Tuesday, 3 June 2008

Issues arising from reading Bundy's 1994 paper

  • Read the other papers from the Behavioral and Brain Sciences volume in 1994. 
  • Read Chapter 3 of Boden, and other chapters, to pin down Boden's definitions of the two concepts (real creativity and mere novelty) as Bundy has cited them. 
  • Check my interpretation of modelling vs simulation... when thinking about this paper's comments, I constantly got confused between the two. I need to clarify this in my head.
  • Follow up citations of Bundy's work, to see if anyone has actually taken these ideas of the need for complexity measuring and self-reflection in creativity further in a practical way (As of the time of writing, haven't found anything yet... possibly i could?)

Bundy: What is the difference between real creativity and mere novelty?

This is a paper published in 1994 as part of a collection of commentaries on Margaret Boden's 'The Creative Mind' book, which had been published 4 years previously. (At some point I really need to sit down with this entire journal section and read through it, when I know the Boden book better...)

Alan Bundy has seized on two distinct concepts from Boden's book, of 'real creativity' and 'mere novelty', and has critically examined both Boden's viewpoint and other ideas on the distinction between these two.

Boden's definitions of the two, as quoted by Bundy, are:

Real creativity: "mapping, exploration and transformation of conceptual space"
Mere novelty: "Generation of new objects from existing conceptual space"

(I haven't found these exact quotes in the Boden book yet so can't add page numbers yet)

From reading the paper,  I think that Bundy's point is that the difference between the two (definition wise) is that real creativity involves some changing of the conceptual space parameters (i.e. the realm of all possible ideas in that domain). So he challenges Boden, saying that real creativity can be present even when new ideas are generated without changing the conceptual space. However I think Boden's definition of real creativity acknowledges this, under the use of the words 'mapping' and exploration'?

This paper does give a very thought-provoking discussion of what is needed to justify labelling something novel as being creative. Bundy suggests that the complexity of new objects being generated is important in how creative they are perceived to be, and that there has to be some kind of self reflection on the part of the creator, to judge the worthiness of such new objects. 

Some points for consideration:
Why isn't mere novelty creative? Is this the right distinction to be making? What is creativity if it isn't just mere novelty? (the paper deliberately only hints at an answer to this question rather than attempting to solve it comprehensively; little has been done to follow up Bundy's suggestions). Assuming creativity must incorporate novelty, what must else it have to distinguish itself as definitely creativity rather than just the generation of novel but uncreative ideas?

I can't find the definition of 'real creativity' that Bundy cites, in Boden's book, although I can find the definition of mere novelty easily enough. I would like to see this definition for myself in Boden's own words.

Interesting sentence: '"Real" intelligence appears to arise from the interplay of a number of relatively mundane processes.' (preceeded by: 'it is usually a mistake to regard one aspect of an intellectual process as the key with the others playing only a supporting role')

This sentence makes me think: is Bundy confusing creativity with intelligence in this paper (deliberately or accidentally) - and actually where is the distinction here, the dividing line?or is creativity a sub-part of intelligence? or is creativity = novelty+intelligence? or less guided? more guided?

(but also this sentence links in with some ideas I have had on creativity emerging from concurrent processes, linked to Koestler's bisociation matrices perhaps....)

Is Bundy presenting creativity as intelligent generation of new concepts/ideas? I think he is. Do I agree with this? What are the alternatives? Seems a logical viewpoint to have.

Thursday, 27 March 2008

Library training

I went to a one-to-one training session in the library today, to look at how best to use the online resources. Having quite a sketchy knowledge of this beforehand, it was good to see what was out there - especially for the music side of things which I am rubbish at finding articles about that are relevant to me!

Main resources I should look at for searching:
Scopus, Web of Science (both general coverage)
RILM (music)
IEEE (engineering-related topics)
PsycInfo (psychology)
and MLA for the current linguistics-related project I am looking at

Also JSTOR and ScienceDirect are fairly useful. Google Scholar isn't too bad but it will link to a lot of articles you can't then find online (something that I have found in the past).

Accessing resources (online and print)
For off-campus access, if you access resources through the Electronic library page on the Sussex website, you can log in and then have access to those resources as if you were on campus (this includes having the 'Find-it at Sussex' option to find copies of articles).

One thing I didn't know is that it is possible for us to go into University of Brighton's libraries as well, to a limited extent. So its always worth checking Brighton's catalogue, if Sussex don't have a book/print journal I want (this may be true for me particularly for linguistics material).

Building searches
A good approach is to try several searches, on keywords such as music* and creativ* (using the wildcards for greater coverage), then combine the searches together to narrow them down, rather than doing a new search each time. If I register on Scopus and WOS, then it will remember my search histories and let me save/export searches. I can export Bibtex style references straight from these databases as well. Should help as I have just set up a single Bibtex .bib page for all my references.

To find good keywords, most papers are tagged with author keywords, database-relevant keywords or descriptors. NB to search for a keyword, this will extend the search to title, abstract, author, descriptor etc, so it is best to search by keywords.

Thats the main points of what I got out of today.

(Another useful tip was with the books I have asked the library about ordering - to actually find out whether they've come in or not, and to reserve them for me when they do come in, I should ask at enquiries to reserve it using a manual form rather than online)

Current priorities

  • Write a 150 word summary of my LCC conference abstract and submit it. By March 31st (show this to Nick before sending).
  • Write the full paper for the CIM conference. By April 30th. I need to get a first draft to Alan (my old supervisor and co-author) ASAP
  • Prepare an abstract for my talk in Edinburgh. By April 4th (use the abstract from the full paper for CIM). Email title and abstract to Karen Ludke.
  • Prepare the Edinburgh talk - need to get my MSc system up and running again and work out how to record it in action? Or find out what facilities are available at Edinburgh for a live demo? (Alan would know)
  • Paperwork for annual review - by 16th May
  • Book travel and accommodation for Greece conference
  • Book train tickets to Edinburgh and check the expenses arrangement with Karen Ludke
  • Work on the linguistic creativity project for LCC conference poster
  • Book travel and accommodation arrangements for ICMC - we're going to go even if we don't get accepted to present a paper!

Phew that little lot should keep me busy!


So I have decided to keep this blog going as a form of recording what I do academically. Lets try it for a bit and see how it goes.

Plan is, that when I read new stuff, I will still carry on making rough notes manually but then I will spend a little time after reading, to try and summarise/review what Ive read about. I want to do the same things for particularly useful lectures and seminars etc as well.

Quick update on the AHRC situation - I got my form in, on the last day, so just in time! It is being processed now and apart from one form which is still being filled out at my old university, Edinburgh, I don't have to worry about it anymore. I won't hear the AHRC decision till late August - it will prob be the last funding application I get feedback from.

But one more tiny bit of funding to apply for - I'm going to see if I can get a travel award from AISB for covering the costs for my attending the conference in Greece or for ICMC. Need to join AISB (but it looks quite useful anyway) and find out a little more, but I've got in touch with someone called Therie from Chichester building who is helpful here.

Thursday, 13 March 2008

nope, it wasn't the final post!

Just wanted to write quickly that:

With the AHRC funding that I am applying for, there is a pre-submission round of selections within the Informatics Dept at Sussex. In fact, there is a cap on the number of applications that Sussex students can submit such that only one student (masters or doctoral) can actually have their application put forward to AHRC for consideration. This is before it even gets considered for funding!

I heard last night that my application for AHRC funding is going to be the application selected by Informatics :D on the proviso that I can get the form back to Informatics in time for it to be processed and sent to Sussex House by 5pm tomorrow. This deadline is something that I think the department itself has just found out about!

At least (with the encouragement of my supervisor to get the form in early) my form is currently in the process of being sent around referees. The strange part of an AHRC application is that I fill in my part of the form, then send a paper and electronic copy to my first referee. Then the first referee adds their reference to the two copies of the form and forwards that onto the second referee. The second referee adds their reference, and then the whole combination is sent onto the department where you are planning to study.

So my overriding concern now is to get that form back from my second referee in Edinburgh - I'm trying to contact him and I'm hoping that he can get it sent on a next-day post... keep your fingers crossed for me!

Tuesday, 11 March 2008

Final post...? Course reflection

This blog has formed part of the assessment for the Advanced Technical Communications course at Sussex university, which I have to take as part of my first year of doctoral study in Informatics. Whether or not I will continue this blog after the course is finished, I am not sure yet. But this will be the final post in the blog that is relevant to the ATC course; it is a reflection on what I feel I have learnt during the course.

There were a number of different modules during the course - I will look at each one individually.
The purpose of this part of the course was to look at the role of blogs in a technical writing scenarios - something I had never encountered before so this was new information to me. I had previously regarded the role of blogs as a personal diary kept online, which may explain my bias towards keeping the blog as a personal record of thoughts and observations as opposed to a medium for communicating with a peer community. I still think this role of a blog has good value, in terms of self-reflection and recording progress. In the future I may keep a blog-style report of my work, of a more technical nature, although I think the suspicious side of my nature will mean I keep particular details only in offline records to which I can restrict the access. Not that I'm paranoid that people will steal my ideas... but a lot of the time, the information that is most useful for me to consult is the way my ideas progress over time. If I'm cautious of recording the exact details of my ideas, or try to present them in a publicly accessible form, I think I lose some of the usefulness in recording my ideas - some of the value that only becomes apparent over time.

Elevator Pitch
Even though I've only been here for two terms, I am already so used to the question 'So what is your research on?' I've stumbled around with various attempts at answers but never really given much thought to how I should answer this. I guess this is because it is usually friends or casual acquaintances that have asked me about my research so far. Now though, I have been thinking about times when I would really need to be able to present a strong view of my research, for example at conferences etc. So it was really quite useful to give this some thought and think of ways in which I could really make people interested.

I have to say that when asked about my DPhil study now, I still automatically say 'well i'm looking at Musical Creativity' or something similar, before using my prepped elevator pitch descriptions, but actually that seems to get people interested anyway as its quite unusual. But I'm trying to get out of that habit as well, as people get a bit confused at first when I talk about musical creativity. Well I guess my elevator pitch is a work in progress!

Technical posters
For this part of the course (and equally so for the paper writing - especially the abstract/title focus - and for the presentation modules) I had an actual project that I was working on (my voice separation project), that I was focussing on presenting to a peer community. This meant that I got real practical value out of these aspects of the course, in a here-and-now basis, as well as preparing for future occasions.

Although I am not presenting a poster for this voice separation project, the exercise of reducing my (then written and submitted) 4 pages of project paper to a poster format was an interesting process. I cut out so much detail from my poster, but was still guided by the text I had already written. On reflection, if I had produced the poster before the paper, I would have a less clear view of what I was trying to achieve, and the important aspects of my project. However I would probably have been able to represent these aspects in a more intuitive (probably more diagrammatic) format, rather than falling back on summarising the text I had already written.

At the point of producing a poster for other projects though, I think I will have a less defined idea of what the poster content covers (as the work will probably not be a finished and evaluated system, but more likely will be reports on a work in progress and initial results and evaluation). So I will be writing from a slightly different viewpoint. When I was creating this poster, I was still in the mindset of expressing the information about this project in a paper format. So the end result is a poster that I would probably need to change before displaying. This is a shame really as I was quite keen to produce a larger poster after the course conference, for display outside our office.

Seeing as I have had a poster proposal accepted for a conference in Brighton this summer :) , though, I will get plenty of chances to try this poster out again. I think in essence I reduced the information in my poster to the right amount, but tried to express it in the proverbial '1000 words' rather than the 'picture' (maybe 1000 words is an exaggeration!). So I should get that poster on my outside wall after all!

Paper writing
I had done a little study before on how to structure papers, so this part of the course was mostly reinforcing what I had already looked at, but the major point of interest for me here was the deep discussion into abstracts and titles.

Writing abstracts has always been a thing of mystery to me... it was great to have a real structure to follow (I work much better this way, I think!) and to see the exact point of the abstract. As a side point, it was interesting to look at abstracts I personally have found helpful or not helpful, and analyse them according to the exact information they present - this was quite illuminating!

I never really give much thought to the titles of what I write, but I did see in ATC how much rests on having a good title (especially looking at the papers listed for a conference in Music Tech, and judging each paper based on the title. It has had some effect - I have submitted two papers this year, and the first title was a bit poorer than the second (which were written before and after ATC, respectively) - not quite so concise and a bit techy.

1. Artificially Intelligent Accompaniment using Hidden Markov Models to Model Musical Structure.

2. Voice separation in polyphonic music: A data-driven approach.

The part where we looked at peer review papers was also fascinating for me, given that I had actual reviews of my first paper (very very short reviews!) and was glad to see that generally you would get much more feedback than I got!

I think I might have missed this part of the course due to writing up my paper for ICMC. Shame, would have been good! But I looked quickly through the slides online before my presentation.

As a side note, I tried something out with my slides for the mini conference - I went to a 'DPhil Skills' workshop presentation about Powerpoint slides which was very dismissive of using bullet points and instead advocated more readable slides (but still not much text - lots of pictures) and also the use of full statements for slide titles instead of short sentence fragments. I thought I'd try this out to some extent in my slides. I didn't get any feedback directly on that (yet...! but maybe I will) but it actually helped me a lot, with the titles, to see exactly the point I was making in that slide, and helped the construction of that slide as well, because everything on that slide had to back up that statement. I think I'll keep that to some extent, although I have to admit that the bullet point style is too easy to sneak back into the slides. I'll compromise slightly there!

Assessment sessions
The mini conference scenario, including the poster session and the series of talks, meant a lot of preparation work that I thankfully could use my ICMC paper for, but I'm glad that I could get the chance to do this mini conference. It was really helpful for simulating how conferences would work, and I feel quite a lot more confident about how things will be in my upcoming conferences now. Especially, the poster session has been demystified somewhat for me.

This course has been a lot more work than I would like it to have been! But it has been very useful, and I have to admit its been more useful than the previous term's 'Research Skills' course, which was also compulsory for first year DPhils. There are a lot of things about life as a researcher which I was very hazy about, such as what exactly posters were for and how conferences worked, and the whole cyclic nature of getting papers published, and I really felt that these were clarified for me. It was a very good atmosphere in which to ask some questions about these as well - in other arenas you almost feel like you should already know all about conferences etc, without having to ask seemingly 'dumb' questions.... but here I never felt like my questions were too 'dumb'. So overall - a bit invasive in terms of work (which is fair enough I guess for a course that also is a master's course so should have a corresponding workload) but the work was very helpful and worth the effort. Cheers Geraldine!

Good technical writing: a critical reflection of what worked well in the paper.

Here I review a paper that I have found very good: informative, with useful and relevant discussion, pertinent examples and a style of writing that conveys the information easily to me.

This paper is by Geraint Wiggins, a major academic influence of mine (and really not based too far away from Sussex as he is in Goldsmiths in London - I should really try and get in contact with him at some point very soon when I have a little more reading under my belt and a little more developed knowledge). The paper was passed onto me by one of the MSc Creative Systems students, and describes an attempt to encode creative systems in a mathematically rigorous way (using set theory).

The paper's title is: Categorising Creative Systems. So for a start, the title is clear, concise and understandable. It is obvious that he is looking to make some effort to formalise the description of creative systems.

A major thing for me is that the paper is short - 4 pages. This is probably because it is a workshop paper published in 2003 at the IJCAI workshop on creative systems (a forerunner to the 2007 workshop I looked at in the last post). Wiggins packs a lot of information into this paper. (He is helped a lot in this because he is building upon a previous publication rather than publishing brand new work; but the paper is easily read as a standalone source of information as well). I have to admit that short papers are far more attractive for me - having read a lot of waffle (including plenty of my own), I am very appreciative of when information is expressed concisely.

(btw I know that this blog is the perfect example of inconcise expression - however I am not too worried about that as this fits in with what I perceive this blog to be: an arena to express my thoughts in words, in a raw form, 'warts and all', rather than a refined and highly considered format.... but at the same time, I apologise to Geraldine and my peer reviewers for the resulting length of verbal output in most of the entries of this blog!)

The three-sentence abstract again is easy to read - Wiggins introduces what he is about to discuss in the paper, what the aim of his work is, and a longer term view of the role of this work. Ideally I would have liked to see a little more detail of the exact work being presented in the paper, but this is only a small consideration for me - my prime concern is that I can understand the full abstract and have an interest in reading the whole paper, which this abstract achieves.

One thing I noticed instantly about this paper was the use of entertaining subtitles. In essence, the work that Wiggins presents is quite dry: using set theory to categorising creative systems. However subtitles such as Hopeless Uninspiration and Productive Aberration are quite interesting to see and definitely capture my interest! Alongside the more traditional titles such as Background and discussion, there is a good balance of standard academic presentation and an interesting style of writing.

In general the structure of this paper is very neat, with small sections, and an easy pathway through the surface of the content is given by just reading the titles and subtitles - I like that a lot as a way of getting a general overview. Wiggins also tells us in the introduction what each section is going to discuss (and each section does then indeed discuss that content!). The conclusion and summary paragraph, although brief, sums up well what the paper has told us and points the reader in the direction of future work, justifying the progression from this 2003 point to the future work suggested.

Looking back into the content of the paper, the language is understandable and just at the right technical level: not patronising but also not overcomplicated. It is easily accessible to those who have not read the previous work that has been cited, or to those who are a little rusty on set theory (me!) or creativity theory. In particular, the example that Wiggins uses to illustrate the abstract theory is very well presented. I possibly would question how he fits this example to Margaret Boden's original theory (which is the theory he is using as a basis of his theoretical framework); however this indicates that I do have a good understanding of the general more abstract notions that he is discussing, and can engage with this discussion as a result, to refine my understanding of smaller details and perhaps notice areas of the theoretical discussion that I would be able to improve on or highlight as needing further work.

Finally the 6 references that Wiggins gives are useful for directing the reader to further work that is relevant. He doesn't include any 'tenuous' references; instead you can see exactly where in the paper each reference makes an important contribution to the content. At the same time he has some important references there such as Margaret Boden's seminal work on creativity (The Creative Mind, 1990).

So in conclusion: the paper is well-structured, concise and expressed at a very good level of technical description. What could have been quite complicated set-theoretical discussion is made accessible to readers such as myself, who have some background in this area and want to learn about the resulting framework rather than spend time understanding lots of theory. This review is carried out after a few skim reads of this paper. I think I will follow up the leads next and make sure that I completely understand the theoretical framework he proposes, but I have a suspicion that this framework (now 5 years old) may have considerably been updated. This paper definitely leaves me wanting to find out more!

Current Community discussions

As I mentioned in a previous post, I have had the proceedings of a recent conference workshop sitting on my desk all term, with no time really to look at them (even though they are very appropriate to where I want my research to go!). So now is a great time to open up that proceedings book, and talk about the themes and issues being brought to the forefront at this event...

4th International Joint Workshop on Computational Creativity, June 2007, London (Goldsmiths)

Even though the requirements for this blog entry say that I should really be discussing a conference, my particular interest in computational Creativity is not quite mainstream enough to attract regular conferences - yet. However this particular workshop is an independent event rather than a conference symposium, running over three days. It comes from a merging of two workshop series, from Creative Systems Workshops and the AISB Symposia on AI and Creativity.

Paper acceptance rate
27 submissions were made to the workshop, reviewed by at least 3 peer reviewers, and this resulted in 17 papers being presented and 3 posters accepted in addition - an acceptance rate of 20/27. Quite high but in this pioneering field a lot of work is being done in early stages and the field has not reached a maturity or depth of research yet which would merit a harsher rejection rate.

People involved and community coverage
The list of reviewers is a comprehensive list of prominent workers in Computational Creativity. Significantly (to me at least), Margaret Boden is not part of the Programme Committee, despite her important contribution to creativity theory and her near location to London. However I understand from recent conversations with researchers here that she is moving away from being involved in creativity and has always held more of a cognitive or philosophical outlook on creativity rather than a computational interest, (despite certain passages in her publishing suggesting the beginnings of a computational framework for creativity). There are a few other people missing who it would have been great to involve, but factors such as location, the lack of perceived prestige and the high specificity of the workshop theme may well explain this. Eduardo Miranda, Alice Eldredge, Philip Johnson-Laird, Robert Sternberg, David Cope and Doug Hofstadter spring to mind here (although some of these people are very high-profile academically and their involvement in a conference would instantly bring the conference much wider international kudos!). However, the presence of people such as Geraint Wiggins and Amilcar Cardoso (co-organisers), Pablo Gervas, Graeme Ritchie and Gerhard Widmer on the program committee means that there are some significant researchers here from this field. In addition, there are submissions from people such as Elaine Chew, Jon McCormack, William Latham, Ana Paiva and Chris Thornton (my supervisor), who are all people I have come across in one way or another during my two terms at Sussex. Papers are contributed from across Europe, from Australia and also from North and Central American universities - so a wide coverage.

(Its interesting that even though my voice separation project earlier this term doesn't immediately spring to mind as an obvious link to computational creativity research, there are a number of people at this conference who have carried out similar work on voice separation: Gerhard Widmer, Elaine Chew and Ana Paiva. Makes me think that maybe I can justify the inclusion of this work into a thesis on computational creativity in music... but those thoughts are definitely for a later date!)

Themes at the workshop:

I'll start with the two themes of most interest to me: Musical Creativity and Frameworks for Creativity. Then I'll work through the other themes in order of interest.

Musical Creativity

This is very interesting to me; as such little work has been done from an academic perspective on computational musical creativity, I am eager to find discussions about it. The papers presented here on musical creativity take diverse approaches, from Artificial life to statistical models. This is great as there are so many possible options here, and the field of research is very young and ripe for exploring. However there are only four papers presented at this session, I would have liked to see more depth (obviously!)

The music that is created in practical systems ranges from chord-based accompaniment for existing songs, to randomly generated musical melodies. It is interesting to see the accompaniment generation work, in particular, as this involves more musical factors than single instrument melodies due to the possibilities in the interactions between each note being played at a single time. But this work looks to be quite musicologically based, in other words it is a result of following analytical rules, rather than looking more at inspirational artistic creativity. I guess I want to see more on generation of completely novel music, without following pre-determined, strict 'rules of thumb'. I might be a little bit ambitious here in what I am looking for though, given the current state of research here.
A related paper of interest here is a discussion of evaluation and systematic improvement methods in models of harmony - definitely worth more than the skim read I have so far given it as evaluation will form a major part of my PhD thesis.

Frameworks for Creativity

As I have a very strong interest in Creativity theory, I am approaching my work from the viewpoint of realising theory in practice, rather than retrospectively engineering creative practice to fit a theoretical framework. So discussions at this session of the workshop are very relevant to me (* although I mustn't neglect research into practical creativity just because it isn't my specialist area...!)

My supervisor, Chris Thornton, presents an interesting view on Margaret Boden's creativity theory as a quantifiably realisable framework rather than a qualitative discussion of differences in creative practice. I have learnt a bit about the work he is doing in this area but it is useful to read papers like this where he puts his thoughts down to paper. Some interesting references here to look up as well.

As this is such a fledgling area of research, the papers here don't go into much depth but instead present brief discussive overviews of computational creativity theory - as I would expect to see. Again there are useful considerations of problem areas and evaluation techniques, so I think that I'll get a lot from analysing the contents of this session in more depth.

Applied Creative Systems

Like I have said earlier, I really should show more interest in this than I currently do. The actual implementation of creativity theory will hopefully for me result in a creative system or systems of my own, so I need to get a more deep view of the creative systems currently being made, rather than the broad but shallow view I currently have. There is some very interesting discussion here of creative systems demonstrating humour (which I have heard of) and one that comes up with advertising slogans (which I wasn't aware of). The creativity theory in the literature has been incorporated into the discussions to some extent, but I would like to really read these papers and see if I can work out exactly how much of the theory has actually been incorporated in the design of these systems. From previous exposure to creative systems, I suspect that only a subsection, if any, of the theory has actually been directly implemented in the systems. I'll see at a later date if my suspicions are well-founded?

Analogy and Language, and Creativity in Narrative

To be honest again, I really don't know much about this area. I am trying to be quite selective about the reading I do here, as it is a little off-topic. However they do propose some interesting modelling attempts here, which is relevant to me as I want to create computational models of creativity. Also I am starting work into models of creativity in language, with the perspective of replicating this work in musical creativity (I'm quite fascinated by the various links between music and language, especially at a neurological level - also linguistic creativity is much more studied than musical creativity, from a computational viewpoint). Pablo Gervas is one of the people I have come across so far in work on linguistic creativity, so I will take a look at the paper he has submitted here, and check out the others as part of the reading for this project, but otherwise I may spend my reading time elsewhere.

Are there any important themes that haven't been discussed at this workshop? I have to say, I can't think of any in particular, although I am not yet very up to date with current research into computational creativity.
I would have liked to see more discussion of the links between theory and practice in creativity, as this is my concentration, but I'm aware that little work has been done here. I'm sure it would have come up in discussions to some degree as both Geraint Wiggins and Pablo Gervas have done work I am aware of in this area, but then without having been to the workshop I wouldn't know (next year though...!)

So all in all, there is some useful content here, even though some people don't appear to be involved in the workshop whom I would like to have seen mentioned. As my supervisor, Chris Thornton, was involved in the workshop, we have spoken since about this workshop. He described it as interesting but not a major signpost in academic advancement! Which seems fair enough - I think I agree.

Monday, 10 March 2008

(not quite the next post) - a little breather!

Just wanted to record in this blog that....

I've finally finished all the funding applications!!!!

well nearly....

All my work on the applications is done now, and I don't think I'll find any more places to apply to (although I need to check out one organisation I found once called SEMPRE...) So the only work I have to do now is to chase up some forms if they don't get forwarded on by my previous universities, and then prepare for any interviews I might get offered. So, for a while...

No more form filling!!! :):)

Who is my peer community?

Looking at the last part of my last entry, I think now is a good point for me to record exactly who the peer community is that I am aiming to become part of.

I think that there are many different layers to this peer community, and the separating criteria between these layers are the different levels to which I want to interact with other academics in this area. Let me explain....

So my work is highly interdisciplinary; I am influenced by (in no particular order):
  • Informatics including Artificial Intelligence and Computer Science
  • Cognitive Science
  • Music analysis, Musicology, Composition and Performance
  • Psychology
  • Philosophy (to a very small extent - because of my lack of knowledge and ability rather than lack of interest!)
  • Maths
  • Linguistics and Natural Language Processing
  • Neuroscience and biologically-inspired/evolutionary methods
I think I could even add to this list, but its long enough already!

So to what extent do I want to become involved with, say, the Maths or Linguistics-based academic community in comparison to the Informatics based academic community? This would be far less than for Informatics, because of demands on time and the amount of information my poor little brain can actually process. However I really don't want to exclude what something like Maths has to offer, not completely. I want to stay aware of relevant strands of information.

So, in all these areas, I am keeping aware of things like departmental seminars and local talks, and occasionally looking at courses within these departments that might be worth sitting in on. However my peer community is, I suppose, concentrated around Artificial Intelligence, Cognitive Science, and particularly both those subjects with a Music bias. (My Dphil study addresses Computational Modelling of Musical Creativity)

Here is a not so little collection of people that I am particularly keen to keep knowledgeable about and maybe get to meet. These lists are the product of a couple of term's work...

People, where they are based and their interests that concern me
Ani Patel (Neurosciences institute: Music and the Brain)
Bryan Pardo (Northwestern university: AI and music)
Chris Darwin (Sussex: Perception inc. sound perception)
Chris Raphael (Indiana: Automatic accompaniment and statistical methods in music)
Diana Deutsch (UCSD: Psychology of music)
Emilios Cambouropoulos (Thessaloniki, Music Informatics) - the co-organiser of the course in Greece that I am going to and a former PhD student of my old supervisor - so hopefully i'll be able to talk to him sometime at the conference)
Geraint Wiggins (Goldsmiths: Computational creativity)
People at IRCAM (Music informatics research institute in France)
Isabelle Peretz (Montreal: Music Neuroscience)
JJ Bharucha (Tufts: Music Neuroscience)
People at Jyvaskyla university in Finland: Music cognition labs
Michael Mozer (Colorado: Neuroscience inc Music)
Maggie Boden (Sussex: Computational Creativity)
Nick Collins (Sussex - my supervisor! - computer music)
Chris Thornton (Sussex - my second supervisor - creativity and machine learning)
Philip Johnson-Laird (Princeton: creativity psychology)
Eduardo Miranda (Plymouth: Evolutionary Music)
Roger Dannenberg (Carnegie Mellon: AI in music)
W. Jay Dowling (Texas: Music Psychology)
David Meredith (Taptu.com and formally Goldsmiths: Music information Retrieval)
David Huron (Ohio: Music Cognition and computational methods in music)
Mihalyi Csikszentmihalyi (not sure where he's based: Creativity theory)

Phew- and that was an edited version of the lists I keep!
So many people whose work I find fascinating... such little time to find their work and get in touch with them. Generally I am hoping to meet people at conferences and also to approach them when I am working on projects similar to them, for example I've been intouch briefly with Chris Raphael about a former project I did in AI accompaniment.

The major conferences of interest to me are:
ICMC - International Computer Music Conference
ISMIR - music information retrieval
MIREX - music information retrieval
ICMPC - Music Perception and Cognition
NIME - New Interfaces for Musical Expression
AISB - simulation of behaviour

How can I get involved in these conferences? Obviously I can't go to all of them! or very many of them in fact. So I need to combine attending conferences with reading the proceedings for others.

So far I have submitted a paper to ICMC - lets see how that goes! and extensively used papers from ICMC, ISMIR and MIREX in various projects up till now. I'd like to find out a little bitmore about the AISB and ICMPC conferences though, especially the recent AISB workshops on computational creativity which are next on my list of things to read about.
There are a few smaller conferences that also interest me, such as CIM (interdisciplinary musicology) and LCC (language communication and cognition) which I will be going to as a presenter of a paper and poster respectively, and also CMMR - Computer Music Modelling and retrieval, which I'd like to look at the papers for when I get some time.

Generally these conferences accept both papers and posters, the academic standards, although many music-oriented conferences take demos of systems in operation or musical performance recordings as well, and have gigs alongside the academic presentations - such as ICMC which should be a very interesting week in Belfast this year! Requirements for submitting vary widely - the three conferences I have submitted to so far have all wanted different things - a full paper, an extended abstract and a short abstract. Criteria for acceptance is standard for academia in Informatics, (I have this impression anyway!), with conferences such as ICMC being of a higher global reputation and correspondingly harder to get into, but conferences such as CIM being smaller (and more focussed on particular themes rather than a general coverage).

Another way I want to keep up to date with current research is by scanning the latest articles published in various journals. CurrentlyI have bookmarks, RSS feeds and emails telling me titles of the latest articles published in: Computer Music Journal, Journal of New Music Research, Leonardo and Leonardo Music Journal, Psychology of Music, Music Perception, and various other Cognitive Science/Neuroscience journals. I try and have a look at these titles fairly regularly but will only click through to a title if it really interests me (because of time restraints). However I quite like the fact that this gives me a wee idea of whats going on, even at the most brief of levels.

A look ahead at a future entry: Current Themes of interest in my community
Current themes of interest particularly in my area: Well the massive one is computational creativity which is a real buzz-word at the moment. Also evolutionary methods are getting very popular (with a very strong bias in this direction at Sussex, although this isn't yet replicated everywhere and I doubt some institutions will consider this approach worthwhile for many years to come - although in such a public arena as a blog I won't go into detail here!)
I'm very interested in the latest advances in neurologically and biologically inspired cognition, especially computational modelling of various levels of cognition. This is getting to be a very exciting area.

In the next post, I'll take a look at one of these themes in particular: a recent discussion on computational creativity. I have had a book of proceedings sitting on my desk for this conference for weeks now, without having looked at it properly. So here's my chance!

OK I've written so much here, but this is an area I've been really looking into over the last two terms, and with so many different academic influences... I know I will have to focus my thesis into smaller areas, but I really like the multidisciplinary ethos at Sussex and how I can incorporate it. This was a major attractor of Sussex for me when choosing where to go for a PhD. So what if there is a greater amount of reading to do, or more areas where I feel like a dunce until I've done some work - I'm really enjoying this research approach! I will focus as time goes on, but not just yet...

Criteria for my blog

I realised that I haven't actually documented anywhere who this blog is aimed at, or how I am going to use it. So here is a retrospective look at how I've been writing the blog content:

Who is the audience
Primarily the audience is mostly intended to be myself; I am treating the blog as a 'dumping ground' for the thoughts and reflections that I would normally record on paper or just in my head. Other people that I think may read the blog are other people in the advanced technical communications course. I doubt anyone else will stumble upon it: in particular although I have told my supervisor about it, I wouldn't think he would have the time to read through it, and I'm not expecting him to. So I am really writing this blog as a personal reflection but making the contents public for the benefit of the ATC course.

What tone will I adopt
In keeping with the above point, I have mostly used quite a conversational and informal tone. If I am mostly communicating my thoughts, then I don't want to rephrase them in more formal language, but to keep them as raw thoughts. I have, in an earlier entry, practised some more formal writing, which was very useful for formulating my abstract for my voice separation paper. But I have quite a few outlets for formal writing this term - papers and funding applications - and wanted to keep this relatively informal.

What types of content will I write about
In addition to the required entries, which I have been considering the content for over the last week, I have used this blog to 'dump' my thoughts and self-reflections/organisations in, as I mention above. I do this to some extent on paper in a notebook I keep on my DPhil, but for this term I decided to try and use the blog for this. It has been relatively useful, however I can't bring this blog to lectures and talks in the same way as I can bring a notebook (not without a laptop anyway).
One way this blog has been useful is that I can always access the content anywhere there is a computer,rather than having to have the notebook physically with me. Of course with this there is the proviso that I need to have a computer nearby, attached to the internet...

To what extent do I want to engage with a peer community
I assume this refers to this blog in particular, i.e. to what extent do I want this blog to be a point of interaction with a peer community. I have little objection to the contents of this blog being in the public domain in general (although I'm not sure I will keep the entries online about where I have put down my reactions to other people's papers - they are by no means a considered and well-founded reaction to the papers' content but just my reactions after a few sweeps of the paper, on what I think is the general gist of the paper).

I don't think that an academic peer community would be particularly interested in this blog's comments, however, as it is written as a personal reflection of what I've been up to this term rather than providing any academic contribution. I have my personal and Sussex website up for providing information about me to the academic peer community I am hoping to be considered part of - hopefully those sites do a much better job of this than this blog does.

The only peer community that might be interested to a small degree in the blog contents are other DPhil students, and ATC students (MSc and DPhil), to compare with their own personal reflections. However, as I say, this blog is very much for personal critical reflection and thoughts.

Friday, 7 March 2008

oh dear....

So I don't think I'm really keeping up with the assignments side of this Advanced Technical Communications course so well.. I thought that the required entries we had to make for this blog were to be done by next Thursday but actually they have to be there ready for review well before then, so that the reviewers can submit their reviews by Thursday.

Sorry Gareth and Maria, I've only done one of the required entries so far... I will get the others done ASAP so you can do your reviews - please be kind :)

Wednesday, 5 March 2008

What makes a good presentation?

Now that I have given my presentation and had the chance to watch some other presentations, I feel quite equipped to tackle this 'required entry'. (Maybe it would have been better if I could have carried out this reflection before my presentation, however time wasn't kind to me at that point..!)

I'm going to start this by thinking about what makes a bad presentation:

- Poor communication skills:
Not addressing the audience directly can make the audience feel like the presenter isn't talking to them!
Mumbling - so you can't hear what the presenter is saying - so what's the point in talking....?
- Inappropriate level of content:
The level of detail needs to be right. Too much detail and the audience will be overwhelmed by information and this might obscure the take-home message you are trying to convey to the audience. Too little detail and the audience will just not appreciate what you are trying to tell them, and the talk won't be informative
Content not directed at audience's general level of knowledge.
Need to make the content relevant and understandable for the audience, otherwise the message won't get through
- An 'uninteresting talk'
This is so important! If you don't make it interesting then people aren't going to listen.

There is a general theme across all of these observations -the point of a talk is to convey information to the listeners. You need to be sure what you want the listeners to get from the talk and that this is a reasonable amount of information that you can get across in the time. Presentation of your material is key to this as well, both visually (slides) and how you speak to the audience.

Having been at a couple of talks yesterday, I would have to add that there seems to be a threshold of time that you can talk to an audience for before they lose concentration - even if the audience is particularly keen to listen!

I've been trying to think of ways to get around this. Perhaps it wouldn't be too bad to disrupt the flow of what you are talking about, in order to give the audience something unexpected that attracts their attention. I suppose this should be relevant to the talk (although I remember a very good lecturer who used to stop half way through a lecture, show a random funny video, to give us a break from the lecture material - that worked well!) Another way around this would be to engage the audience in active participation, or to use sound or video as an alternative method of presenting information. I'd quite like to think of other ways around this as well - it would make my presentation style much better, I think, and more individual.

Tuesday, 4 March 2008

demands on time

Oh dear I haven't done any of the posts required for this blog yet, and I'm so busy with funding applications, marking and work for conferences. I _will_ make time for them in this next week - I have to!

This term has been so busy, what with getting so many deadlines met for various things. Especially now I have my list of possible things to be working on, I am realising that I spend far too much time thinking about exactly how to do things rather than getting stuck in. There's an importance in thinking about things, of course, but at the end of the day the work has to be done!

In that frame of mind, then, lets finish the second draft of my AHRC funding application! Last application form I do (I hope!) and then I can get on with the reading I want to do

reflection on presentation

On Friday, I gave a presentation on my Bach fugue project to the members of the Advanced Technical Communications course.

I found this presentation quite tricky to prepare, as in ten minutes I had to explain the problem I was addressing (quite a specialised musicological problem) and the solution I implemented - I left out so much detail that I would have loved to put in! but had to be ruthless.

I think the presentation went quite well overall. I was very worried about the fact that I wanted to use sound in my presentation, but although I got there quite early to try out several versions of my presentation (I nearly resorted to using the backup of the backup of the backup of the backup plan!) I still kept people waiting a little at the beginning of the session while I got Eric to help me set the sound up on the PC we were using. Lessons learnt - its not enough to be there early - if I don't know how to do something, I need to be much more forceful in asking for help setting things up, and not be afraid to admit that I don't know how to fix things.

Overall I think the presentation went ok, and I think I managed to convey what I was trying to do, and how successful my work had been to that point. There are always points to improve upon, and I picked up a few tips watching others - Geoff's air of complete knowledge and authority in what he was talking about, Maria's engaging with the audience during the talk by asking them questions, Chris's demo half way through his talk which was short but exciting to watch and the way that Gareth looked around the whole audience while he talked, not just a random focal point in the group, and how he talked very deliberately and a little slower than he normally talked.

I've enrolled for an Oral presentation skills course which is on next week - I wish it had been on before this presentation! But seeing as I have a couple of talks coming up to give at Edinburgh uni and the conference in Greece (and perhaps one at Warwick uni), I think I'm going to have plenty of opportunity for practice...

Wednesday, 20 February 2008

Ongoing projects

One of my worries before starting this DPhil course was that I wouldn't have much variety of work to be doing, but would instead be spending all my time reading and trying to find suitable material to read.

In actual fact, it has turned out that I seem to have a million and one different projects on the go, to the point that I am finding it hard to keep track of everything. The reading pile is definitely there on my desk, waiting for me to pile through more of it. In addition to this though, I have papers to write, funding applications to prepare, courses to sit in on, interesting seminars to attend, talks to plan and give, coursework for this ATC course (including some specific blog entries that I really should get around to soon) and then the actual end of year report will start to loom, where I am supposed to finally know what I am going to be doing in my DPhil research...

So to try and get a handle on what I'm currently working on, I am going to make a list here, for my own reference.

I have submitted applications to a couple of charitable funds now. So the main funding application I am working on is to the AHRC. Because my work is so multi-disciplinary (computational models of musical creativity), I am also hoping to be considered for EPSRC funding through the department (which I am not hopeful of getting, as Music Informatics already has one departmentally-funded student). There is also a very tenuous case for submitting an application to the Open competition in ESRC, although I think this may not be really worth pursuing.

I will never have time to do all the reading I want to do! But this is a collection of what I would like to at least skim through:

A number of papers, particularly I need to look through the publications list of Wiggins, Cambouropoulos, Patel, Bharucha, Mozer, Johnson-Laird, Miranda, Dowling and Huron, to name but a few!

Books (a small selection, in no particular order):
- Margaret Boden's Creative Mind and Dimensions of Creativity
- Sternberg: The Handbook of Creativity
- Lerdahl and Jackendoff: A Generative Theory of Tonal music
- Sloboda: A musical mind
- Bregman: Auditory Scene Analysis
- Roads: The computer music tutorial
- Koestler: The Act of Creation
- Csikszentmihalyi (sp??) Creativity
- Wiggins/Deliege ed: Musical Creativity in theory and practice
- Peretz/Zatorre ed: Cognitive neuroscience of music
- Todd/Loy ed: Music and Connectionism
- Cope: Can't remember the name of the book but it is about his creative music system EMI

Conference proceedings: AISB 2002, ESCOM ?, a workshop on creativity in Edinburgh in the 90s and workshop in computational creativity 2007

I need to turn my MSc thesis into a paper that matches the extended abstract I submitted for an Interdisciplinary Musicology Conference in Greece, which I will be going to this summer. As I am writing this paper jointly with my old supervisor in Edinburgh, we need to work out how we are going to co-author the paper - I think the ideal scenario for both of us is that I write the bulk of the copy, with a lot of editing input from my supervisor, but I need to check this with him.

Also, I have jointly (with my current supervisor) submitted an abstract to a language and cognition conference in Brighton this summer. It is touch and go whether that will get accepted, I think, as it is slightly off topic for the conference; we are applying linguistic models of creativity to music, acknowledging the parallels between music and language, but the focus of this conference is on language. However this is work that I really want to investigate further, so I am going to carry on with the work anyway. If it doesn't get accepted at Brighton, then it will be useful for my DPhil anyway and might be publishable elsewhere.

Final paper is the one I submitted to ICMC earlier this month. Should it get accepted for ICMC, I am sure there will be some revision I will need to do...

The technical communications course that this blog has been created for is the only course that requires me to do coursework outside of my DPhil study. Although to be honest it is a little frustrating to have to think about deadlines for this course when I am more concerned about the fact that I still don't know what my research questions are going to be for my DPhil, the course is proving very useful and will be worth the effort I think. I have to prepare a poster and a talk for a mock-conference in March, for which I will use my ICMC project. I'm hoping to reuse the talk in ICMC (if accepted) and possibly at my old university, Warwick, as a researcher there is interested in my work and would like me to go and speak there.

There are so many seminars on at the moment that I would love to go and sit in on, but I have so little time. But I am keeping my eyes open. There is one in E-intentionality coming up on the link between creativity and novelty, which I will definitely go to, and a few others that I will try and get to as well.

At the moment, there are courses in Generative Creativity, Data Mining, Music Analysis, Neural Networks and Computational Music, for which I am trying to sit in on at least some of the lectures for (it depends on what that specific lecture is about as to whether I go to it or not).

Actually these lectures are serving a double purpose for me at the moment; seeing as three of them start at 9am and one at 10am, they are giving me a good reason to get myself out of bed and into campus at the beginning of the day! Then after the lecture I feel like my brain has been woken up, and I'm ready to get going with whatever else the day has in store for me.

Sussex offers a lot of training courses at the moment, currently I am doing the Profolio course for first year DPhils which is a few hours every few weeks, and also some skills courses offered by SP2, the training skills programme at Sussex. Seems a shame to waste these opportunities, and they are all helpful in some ways, some more than others.

As I mentioned, I am in the middle of arranging to give a talk at Warwick. Edinburgh have also booked me to give a talk at the end of April, on the content that I am submitting to the Greek conference, based on my MSc thesis (artificially intelligent piano accompaniment). That should do me for now...!

No teaching duties for me this term, although last term a 2 hours/week tutorial job took up a lot more of my time than anticipated. I do however have commitments in terms of marking, which is averaging about 5 hours a time, about every 2 weeks, and some monitoring work for the department which is half an hour a week.

This comes at the end of the list, even though it is the most important for me. At some point I need to finalise exactly what I will be doing the next year. Currently I know that I want to make models of some creative processes in music. So the lecture courses will help me to identify what models I could use and what history of creativity theory already exists, as will my reading. I need to pin down some specifics now.

Phew. That took a while to write down, and I've only included the academic projects I have ongoing and not the out-of-work projects such as music rehearsals, getting more gigs, getting to know more people in Brighton generally, family issues and running London marathon, all of which are time-consuming but important to me. I think its been worth the time though, its good to have a list down in black and white, that encapsulates what I am doing at the moment.
I'd better get on with it now then!

Tuesday, 12 February 2008

Trip to Greece

Following on from the last post...

Got some good news about an abstract I tentatively submitted to a conference in Thessaloniki for July, based on my master's thesis work. The conference is called 'Conference in Interdisciplinary Musicology' and it's run by someone whose publications interest me a lot, Emilios Cambouropoulos. The submitted was on an artificially intelligent musical accompaniment system, that I worked on as my master's project last year.

I co-authored an abstract for this with my old supervisor at Edinburgh (where I did my degree), and our abstract got accepted for the conference. So I'm off to Greece in the summer!

Don't think it hurt our chances of acceptance in the conference, that my old supervisor also supervised Emilios Cambouropoulos's PhD thesis at Edinburgh about ten years ago... but also the review comments were interesting to read. Two reviews were given, based on academic merit, relevance to the conference and relevance to the conference theme (musical structure). One reviewer thought our work was generally ok to good, based on the good application of computational methods to a musical problem but questioning the contribution back to music theory (which is a weak point for the work for this particular conference - I'll have to think about this a bit more). However the other reviewer loved the work, rating it as excellent for all the review criteria. That was very satisfying to read! I wonder if they would have still rated it as excellent for academic criteria if they had known it was sourced from a master's thesis work... :)

Mission complete

Thoughts collected, ideas collated, diagrams drawn, paper written.

It feels pretty good to look back on the paper I've produced on this music analysis project. Even if the paper doesn't get accepted for the conference I am aiming it for (ICMC - computer music conference) then I still have a 4 page paper which describes what work I have managed to do on this project in the last few months. It has pretty pictures and music notation and everything
... :)

At first I wasn't that motivated to get this music analysis project working properly. It didn't quite fit in with the 'music creativity' research I was really wanting to read about. Also, I am a bit nervous about the whole concept of presenting a paper on this at ICMC, which is a big conference for me. I've only been looking into this for a few months, and not exactly a solid few months at that, so to present that work to a bunch of academics who really know their stuff... quite a daunting experience. So if I didn't finish the paper in time, then - oh well!

One deadline extension later, though, and I suddenly had a couple of weeks more than I thought to really get my project working and to understand what I was bringing to this area of research. Having put my ideas onto how to analyse these Bach fugues and separate them out into their different voices, I got something working in Matlab. Didn't work 100%, but to see something kinda sorta working was a massive incentive to carry on and refine the work - to the point where I had to define for myself a point at which I would stop trying to make more and more improvements and instead knuckle down to writing the work up.

The researcher's nightmare of finding some related papers quite late on in the project work happened to me here. This was a real blow to me at first, as I thought: 'oh no, all my beautiful ideas have been done before'! But I started noticing aspects of my work that were really different to the work already done. I guess it was at this point that I realised that even though I'd only been tackling this fugue analysis problem for a few months, I still had valid ideas to bring to this problem and I had proposed a solution that had significant aspects to it, which hadn't been done before. At least I hope they haven't been done before... I've followed up the research trail a little bit more and haven't found anything yet...

So now what happens? Well I have now got a four-page paper under my belt; even if it doesn't get accepted for ICMC then I think I will try and get it published somewhere else (taking any review comments into consideration) as I think I've managed to get some good results out of this work, and I want to get it into the academic domain somehow. Let's see what ICMC say anyway, and go from there.

Tuesday, 5 February 2008

that'll do...

Ok so I think I've identified where my work on voice segregation fits in with the related work that has been published recently. There has been a lot of good progress in tackling this problem in the last five years or so, but it is by no means a solved problem and there is a lot of work still to do. My machine learning/data mining approach does not encode a lot of human knowledge about voice-leading music separation, but performs well at the voice separation task due to what it has learnt during training. Now to write up my work for ICMC and see if people agree that this work makes a good contribution

Re: Bach fugue project. Kirlin and Utgoff 2005: Learning to segregate voices

As the only paper I have found which incorporates some form of learning how to segregate voices by training on examples, rather than by following rules, this is the most similar to my work. However the authors only train their system on very small amounts of data; they use one piece (Bach's Ciaccona) for both training and testing, choosing small sections (circa 4-8 bars) from this piece to train the system on. The training sections are selected due to their similarity to the testing sections chosen from the piece. In the later experimentation there is some effort to combine the earlier training sections into one large training section. However this is still only using a relatively small amount of training music and still uses the same piece, so therefore there is a constant underlying tonality, style, set of basic thematic ideas). Some overfitting may well occur to that one specific piece (Ciaccona). How would this generalise to other similar works by Bach? or other Baroque pieces? or more widely across the musical spectrum? Could their system cope with larger amounts of training data and if so, why haven't they used more?

It is good to see in this paper that they say that they have found no other voice segregation methods that used automatic learning techniques - neither have I (yet)

This method learns how to identify what voice a note belongs to by observing its pitch relationship with the note in the previous time-slot. They examine the piece in smaller windows rather than in one large set, using a fixed window size (the exact window size is varied across experimentation to see which gives the best results).

Also: this system uses decision trees, therefore it is always a single, discrete answer that is given to the question "does this note belong to voice v?". No measure of the likelihood of this answer being correct, even in a non-probabilistic, general estimate. What happens in this system if there is a clash of two simultaneous notes being assigned to the same (monophonic) voice. The authors are unclear on this prospect.

Thinking about Cambouropoulos's points (Voice identification 2006), it is interesting to see that this system makes no use of "vertical" information in its process (harmonic structure at any particular time-point c.f. Schenkerian analysis middle ground analysis level). So they do not make use of observations about what notes are sounding at the same time to guide their working at all.

Another difference between this system and mine is in the end-goal of the system. My focus is on identifying the route that each voice takes throughout the entire course of the piece, assuming that voices are present (but not necessarily active) for the entirety of the music. K+U , on the other hand, take a lower-level approach, identifying fragments of the voices on selected bars, but allowing the voices to vary throughout the course of the piece.

Re: Bach fugue project. Madsen and Widmer 06: Separating voices in MIDI

This method works from left to right, from the beginning of the piece to the end. A cost function is used which is based on the difference in pitch between consecutive notes in a voice: trying to minimise this difference (using Euclidean distances). So there are different levels of correctness assigned to each voice allocation, rather than a definite discrete classification. This would be useful in reconciling cases where there is a clash in assigning simultaneous notes to the same voice (assuming monophonicity of voices).

This cost function is based on predetermined rules inspired by Temperley's preference rules for voice-leading harmonic structure. These rules are an example of how the implementors add human knowledge into the system for it to work correctly. Madsen and Widmer have experimented manually with the setting of this cost function, using their musical knowledge and observations of the performance of the system to improve voice identification success.

Although some human-driven bias is inevitable when capturing a real-world system in a computational model, due to decisions made when structuring the model, I would prefer to keep the influence of human knowledge to a minimum. Instead it should learn as much as possible about how to construct the individual voices from the notes in the piece, rather than use human-devised rules to guide it and later correct it.

Madsen and Widmer's method is to heavily segment the voices so that small patterns of notes (c.f. Chew and Wu's contigs) are detected. They make a small post-processing effort to join these voice segments together but in general their main focus is on local voicings of notes rather than identifying a global voice that is present throughout the piece.

They only test on Bach fugues. I would like to see how their work generalises to other voice-driven music such as quartets or orchestral works.

Monday, 4 February 2008

Re: Bach fugue project. Szeto and Wong: A Stream Segregation Algorithm for Polyphonic Music Databases

A plethora of patterns are perceived in music but the only patterns that are noticeable to us are those that are all contained within a single stream (''grouping of musical notes''). So the authors want to separate polyphonic music into streams, for music information retrieval systems such as indexing a database of musical pieces. Their use of stream seems to be exactly the same as what I refer to as a voice.

There is a useful discussion of the psychology behind why we perceive music patterns monophonically - i.e. one note at a time rather than a chord.

The way they separate music is by identifying clusters of notes that together represent a musical event. Although I understand how they differentiate between simultaneous events (events happening at the same time) and sequential events (events happening at completely separate points in time, with no temporal cross over) I can't quite see how they then link the different events together to form a stream. I think I need to skim read this paper again to make sure I understand it.

Re: Bach fugue project. Chew and Wu: Contig Mapping paper

Very similar work to what I want to do. Their key assumption is that voices don't cross over each other. They use this to allocate voices at points where all voices are sounding.

I disagree with their key assumption! But are they right? Seems a dangerous assumption on which to found their entire work on...

Surely a major problem can occur if you base your whole program on a key assumption from your own domain knowledge, which then turns out to be incorrect some of the time? Better to base the program on objective facts and observations?

But this approach is quite useful in that they cut down the amount of fugue they are looking at, significantly, by using marker points in time at which they are sure that they have successfully allocated notes to voices. Is there some key assumption that I can make that I can use in a similar way? For example if all voices are present, and the notes are very far apart (how would I quantify this) and in the correct range for each voice (ditto)

Their algorithm seems very efficient - O(n^2)

They make several mention of rules - to what extent is their approach rule based?

One major thing I learnt from this paper: if using unfamiliar jargon in a paper, define it early! They don't define 'contig' till p7/20!

writing up a term's worth of work

Not the most normal first post for a blog I guess... but the main purpose of this blog existing is to help me collect my thoughts together in my doctoral research. With that in mind then, the next few posts in this blog are going to help me collect together my thoughts on a project I'm working on.

I'm keeping details of the work I'm doing in a separate diary, but (stupidly) haven't kept much record of the papers I read in the build-up to carrying out this project.

So - I'm going to write short summaries of some papers relating to this project, that will include my reactions to them and how the work in the paper is relevant to my project.

A good starting point for this is to define the problem I am currently working on. Let's see if I can describe it (hopefully this will form part of the introduction for the paper I am writing):

Musical pieces can be made up of several melodic lines interwoven together. These melodic lines are commonly known as voices, although they are not used only in vocal music but are also to be found in music written for instruments. What is important about the voices is that each voice can be considered as a standalone melodic pattern, complete and interesting in its own right. Several related voices, combined together to form one piece of polyphonic music, can generate additional harmonic qualities to enhance the voices.

Fugues are a perfect example of this compositional technique in action, being constructed solely of a number of different melodic voices. J. S. Bach was a fundamentally important composer in the history of fugue composition; in particular his highly influential work The Well-Tempered Clavier comprises 48 fugues. Bach wrote two fugues in every key, to illustrate the full potential of each harmonic key in comparison to each other.

In analysis on Bach fugues, the musicologist identifies each individual voice first; this enables them to perform more advanced analysis on the melodic content such as the re-use of a melodic pattern in different voices. The musical score usually gives the musicologist much help in identifying each voice, as each voice is notated slightly differently (the direction of the note stems indicates which voice each note belongs to).

Identifying each voice would be a considerably harder task, however, if the notational clues were removed. The musicologist would rely on much information within the piece, such as the pitch of the notes and the rhythmical structure. They would use this information in conjunction with their knowledge of how Bach typically structured fugal voices.

Can a computer learn how to perform the same task of extracting the constituent voices from a Bach fugue? I suggest that, given minimal (if any) human assistance and a training set of fugues with the fugal voices already identified, patterns of vocal movement can be identified and learnt. Such patterns can then be used as background knowledge of Bach's fugal voice-writing, to assist the computer in identifying individual voices in a previously unseen Bach fugue.
I'm sure that introduction to the problem area will get extensively re-written but currently I am at the stage of a first draft of a paper describing my solution to this task.