News and Events

Seminar by Andruid Kerne on Computing for Creativity and Expression

Title: Human-Centered Computing for Creativity and Expression
Date: Thursday, Nov 10, 2011, 3:30pm
Location: room 52.321, Roc Boronat building

Abstract: The Interface Ecology Lab develops computing as a medium of expression, coordination, and ideation. The interface-as-ecosystem is a border zone that juxtaposes disparate representational systems. We synthesize diverse fields to investigate -- art, design, cognitive psychology, ethnography, and cultural studies -- with computer science and engineering. We develop human-centered computing for creativity and expression: interaction and information visualization techniques, creativity support tools, games, metadata semantics, programming languages, cultural theories, and evaluation methodologies.
The IEL fosters integrative research projects that span hardware and software. ZeroTouch is a high-resolution multi-finger sensor for free-air interaction or to augment LCD, pen-based, and haptic displays with multi-touch. Trans-surface interaction techniques connect personal and social multi-touch surfaces. Zero-fidelity simulation games, based on fire emergency response practice, teach team coordination in motivating, fun environments. Information composition is a holistic, integrative representation for information collections connecting visual semantic clippings and annotations as a connected whole. Information-based ideation is an evaluation methodology deriving ideation metrics to validate creativity support tools. Support for Information Mapping in Programming Languages (S.IM.PL) constitutes an open-source cross-language type system to support practical distributed computing, connecting iOS, Java, and .NET. Meta-metadata extends the cross-language type system to develop an outrageous alternative to RDF for representing a metadata semantics web.
We observed that these metrics clearly demonstrate an increased level of confusion and unsuccessful search sessions among children. We also found a clear relation between the reading level of the clicked pages and the demographics characteristics of the users such as age and average educational attainment of the zone in which the user is located.

Biography: Andruid Kerne is a researcher working at the intersection of arts and sciences. He is associate professor of Computer Science and Engineering at Texas A&M University, and director of the Interface Ecology Lab. Andruid holds a B.A. in applied mathematics / electronic media from Harvard, an M.A. in music composition from Wesleyan, and a Ph.D. in computer science from NYU.
Kerne's output has been presented by the Guggenheim Museum (New York), ACM CHI, SIGGRAPH, JCDL, Multimedia, CIKM, Creativity and Cognition, TEI, and Document Engineering, ISEA (Paris, San Jose), New York Digital Salon (New York, Spain, London, Beijing), the Milia New Talent Competition (Cannes), the Ars Electronica Center (Linz), the Boston Cyber Arts Festival, the Pan-African Theater Festival (Ghana), and the town square of the village of Anyako (Ghana). His work has been supported by the National Science Foundation, the Rockefeller Foundation, Dance Theater Workshop, the Spaulding-Potter Fund for Innovative Education, and the Texas A&M Department of Computer Science, Arts Foundation, and Humanities Informatics Initiative. He has served as co-chair of the ACM Multimedia Interactive Art program for 3 years, and on program committees including CHI, JCDL, C&C, TEI, WWW, IUI, and DocEng. Press coverage includes Time, MSNBC, Discovery News, Popular Science, PC World, New Scientist, Slashdot, Engadget, Gizmodo, InRumor.com, and Le Monde.

8 Nov 2011 - 12:52 | view
Xavier Serra gives a keynote at ECSS 2011

Xavier Serra has been invited to give a keynote speech at the 7th European Computer Science Summit that takes place from the 7th to the 9th of November 2011 in Milan, Italy. The talk is on "The need of a multicultural approach in IT research".

The ECSS 2011 is the 7th Summit of deans, department chairs and senior faculty of leading European Computer Science and Informatics faculties, departments, research institutes and laboratories. Like its previous editions, ECSS 2011 is the meeting place for anyone interested in issues of research, education and policy in Informatics.

3 Nov 2011 - 23:29 | view
Participation to ISMIR 2011

Joan Serrà, Gopal Koduri, Ferdinand Fuhrmann, Dmitry Bogdanov and Xavier Serra participate at the 12th International Society for Music Information Retrieval Conference that takes place in Miami, Florida (USA) on October 24–28, 2011. MTGs participation includes 2 oral presentations, 3 posters, 3 demos, 1 tutorial and 4 successful entries to MIREX.

Oral Presentations:

Poster Presentations:

Demos:

Tutorial:

MIREX:

21 Oct 2011 - 19:10 | view
First Spanish voices for Vocaloid3 (Bruno & Clara)

On October 21 2011 the first Spanish voices for Vocaloid3 (Bruno & Clara) are presented in the launching event of Vocaloid3 in Japan. These two voices will be commercialized in December through Voctro Labs.

Voctro Labs is an initiative of a team of MTG researchers (Jordi Bonada, Merlijn Blaauw, Jordi Janer and Oscar Mayor). This spin-off company will specialize on voice technologies for the entertainment industry. Voctro Labs was awarded with the second prize in the VALORTEC entrepreneurship contest last April organized by ACC1O.

21 Oct 2011 - 17:51 | view
Kick-off of MIReS project

On the 20th and 21st of October 2011 there is the kick-off meeting of the project MIReS at the UPF. The MIReS project is a Coordination and support action funded by the 7th Framework Programme of the European Commission and coordinated by the Music Technology Group in collaboration with Stromatolite . MIReS will last 18 months starting the first of October 2011 and the partner institutions involved are:

20 Oct 2011 - 19:43 | view
Seminars by Geoffroy Peeters and Petri Toiviainen

On Wednesday, 19th November, at 18:00h in room 52.412, Geoffroy Peeters from IRCAM and Petri Toiviainen from University of Jyväskylä will give two seminars.

18h: Geoffroy Peeters on "Recent Advances in Music Information Retrieval at IRCAM"
19h: Petri Toiviainen on "Acoustic, behavioural and neural correlates of polyphonic timbre"

Abstract Geoffroy Peeters: In this talk, we review recent advances in  Music Information Retrieval at IRCAM. We focus on the "copy and scale" method which aims at replacing signal-processing algorithm for the estimation of local features; the use of Universal-Background-Model for music classification and similarity and the description of the Production Effects of recorded music.

Abstract Petri Toiviainen: Polyphonic timbre is has perceptual significance in, for instance, the identification of genre, artist, and  musical affect. It also has significance for computational categorization of music. Its perceptual constituents have however so far received little attention. I will review the work performed in our lab on the perception of polyphonic timbre. This includes determining the dimensionality and configuration of perceptual polyphonic timbre space  as well as the acoustical correlates of the perceptual dimensions of polyphonic timbre. A further question that we have tackled is the effect
of enculturation and musical training on the aforementioned issues. Moreover, I will discuss our research on neural correlates of polyphonic timbre and other musical elements. Our approach utilizes a novel paradigm, in which real music is presented to participants while their brain activity is scanned using functional magnetic resonance imaging. Subsequently, the thus obtained dynamic brain activity patterns are correlated with musical features that are computationally extracted from the stimulus to identify brain areas involved in the processing of these features.

17 Oct 2011 - 14:35 | view
Phonos: Octophonic Concert
On Tuesday October 18th 2011 at 19:30h in the Espai Polivalent of the Communication Campus of the UPF, Phonos is organizing an Octophonic concert of electroacoustic music.
16 Oct 2011 - 21:30 | view
Cyril Laurier defends his PhD thesis on October 19th

Cyril Laurier defends his PhD thesis entitled "Automatic Classification of Musical Mood by Content Based Analysis" on October 19th 2011 at 13:00h in room 55.309 of UPF's Communication Campus.

The members of the jury are:  Petri Toivainen (University of Jyväskylä), Geoffroy Peeters (IRCAM), Emilia Gomez (UPF).

Abstract: Digital music is becoming a major part of the user experience with computers and mobile devices. Automatically organizing this content is a major challenge. In this work, we focus on automatically classifying music by mood. For this purpose, we propose computational models using information extracted from the audio signal. The foundations of such algorithms are based on techniques from the fields of signal processing, machine learning and information retrieval. First, by studying the tagging behavior of a music social network with dimensionality reduction techniques, we find a relevant representation model for mood. We believe that this methodology can be applied to other domains. Then, we propose a method for automatic music mood classification and detail the results for different types of classifiers. We analyze the contributions of audio descriptors and how their values are related to the observed mood, trying to find explanation from psychology and musicology. We also propose a multimodal version of our algorithm using lyrics information, contributing to the field of text retrieval with a new model based on key words differentiating categories. Moreover, after showing the relation between mood and genre, we present a new approach using automatic music genre classification. We demonstrate that genre-based mood classifiers give higher accuracies than standard models. All the proposed algorithms are evaluated with user data. Our audio based approaches, adapted to the context, have been evaluated in international evaluation campaigns. Nevertheless, we propose a rule extraction technique to explicit the strategy behind our models. This method allows to make sense of the classifiers and to understand how they can predict the musical mood.

12 Oct 2011 - 11:59 | view
Talks on MTG's SIEMPRE work at McGill

Esteban Maestre, Panos Papiotis and Marco Marchini will give two lectures on October 12th 2011 at McGill University, Montreal, related with the ongoing work of the SIEMPRE EU FET-Open project in the "Workshop on motion capture of string quartet performances". Tha talks are organized by CIRMMT. The lectures are:

  • "Towards a computational analysis of inter-dependence in string quartet performance"
  • "RepoVizz: a multimodal on-line database and browsing tool for music performance research"
11 Oct 2011 - 09:34 | view
Seminar by Waldo Nogueira on cochlear implants

On Thursday Oct. 13th 2011 at 15:30h in room 52.321, Waldo Nogueira, postdoc at the MTG, will give a research seminar on "Signal Processing and Sound Perception with Cochlear Implants"

Abstract: Cochlear Implants are implantable medical devices that can restore hearing to people with profound hearing losss. Speech recognition in noise and music perception with these devices are still severely limited. A factor that determines hearing performance with these devices is the signal processing method or strategy implemented in the cochlear implant sound processor. This seminar will present the cochlear implant system and it will review how hearing is perceived with these devices. The most common signal processing methods as well as new techniques to improve pitch perception, music perception and speech intelligibility will presented. The evaluation methods to assess performance with these devices will be also discussed.

Biography: Waldo Nogueira received his Ing. degree (M. Sc.) in Telecommunication Engineering from the Technical University of Catalonia in 2003. He wrote his diploma thesis "Envelope Estimation of Audio Signals based on Predictive Techniques on the Frequency Domain" as an integrated exchange student at the Leibniz University of Hanover. This institution is an active center in multimedia research (audio and video coding and streaming, 3D modeling, face animation, and computer-human interfaces). After completing his Master's Thesis, he worked toward his PhD degree at the Laboratory of Information Technology of the University of Hanover. During his PhD he developed several strategies for cochlear implants, some of which have been commercialized, like the MP3000 by Cochlear Corporation. Furthermore, he did extensive work on designing evaluation methods to assess sound performance using cochlear implants based on auditory nerve models and automatic speech recognizers. This work finished with the publication of his Dissertation: "Design and Evaluation of Signal Processing Strategies for Cochlear Implants based on Psychoacoustic Masking and Current Steering".  In 2008 he joined the European R&D centre of Advanced Bionics in Belgium as a Research Engineer. His main tasks consisted of developing new algorithms for stimulation of cochlear implants and research on electrophysiology of the auditory nerve. In 2009 he became Principal Research Engineer with emphasis in signal processing at the European Centre of the same company in Hanover (Germany). His main tasks consisted of research and development of new audio signal processing algorithms for cochlear implants, electrophysiology and management of research projects with external R&D centres. Since Semptember he is working at the Music Technology Group of the Pompeu Fabra starting a research line on music signal processing and cochlear implants.

10 Oct 2011 - 13:27 | view
intranet