Jan, 2020:  SOKO JKU: Philipp Schwarz was awarded the Adolf Adam Computer Science Prize for his program, which can read usable profiles even from badly preserved fingerprints. More info: bit.ly/36TlCoz, opens an external URL in a new window
 

Nov, 2019: Best Paper Award at ISMIR2019 goes to a paper by our PhD student Stefan Lattner (Sony CSL, Paris) and Andreas Arzt (PostDoc at our institute, now with Apple, Hamburg). Congratulations!
 

Nov, 2019: Invited talk by Hamid Eghbal-zadeh at the Intelligent Autonomous Systems Group of the Computer Science Department of the Technische Universitaet Darmstadt:  Representation Learning under Incomplete and Imprecise Information Conditions.
 

Okt, 2019: Judges' award at the DCASE2019: The CP Team, Paul Primus, opens in new window and David Eitelsebner, won the Judges' award at the DCASE2019 challenge, opens an external URL in a new window Task1 for "Acoustic Scene Classification with Mismatched Recording Devices"., opens a file
 

Okt, 2019: Best student paper award at the DCASE2019: The Team of CP, Paul Primus, opens in new window, Hamid Eghbal-zadeh, opens in new window, David Eitelsebner, Khaled Koutini, opens in new window, Andreas Arzt, opens in new window, Gerhard Widmer, opens in new window, won Best student paper award at the DCASE2019, opens an external URL in a new window workshop for the paper "Exploiting Parallel Audio Recordings to Enforce Device Invariance in CNN-based Acoustic Scene Classification"., opens a file
 

Okt, 2019: Markus Schedl gives an interview in the dorf-tv series "Hörstadtgespräche", on the topics of Social Media Mining, Selection of Music, Listener Profiles, and Music Recommendation.
Listening only to music that Spotify suggests - is that our future? Will we then notice that we only listen to what we already know because
we like it? These and other questions are discussed in the interview with Prof. Markus Schedl, who researches music listener profiles and music recommendation systems. The video is in German. Video, opens an external URL in a new window
 

Okt, 2019:  Explainable Models and Their Application in Music Emotion Recognition: ARI guest talk by Verena Haunschmid and Shreyan Chowdhury at the Acoustics Research Institute of the Austrian Academy of Sciences, Vienna. View more, opens an external URL in a new window
 

Okt, 2019: Gerhard Widmer to give a Keynote Presentation at the Augmenting Performance Conference. Much of current research in Artificial Intelligence and Music, and particularly in the field of Music Information Retrieval (MIR), focuses on algorithms that interpret musical signals and recognise musically relevant objects and patterns at various levels – from notes to beats and rhythm, to melodic and harmonic patterns and higher-level structure -, with the goal of supporting novel applications in the digital music world. This presentation will give the audience a glimpse of what computational music perception systems can currently do with music, and what this is good for. However, we will also find that while some of these capabilities are quite impressive, they are still far from showing (or requiring) a deeper „understanding” of music. An ongoing project will be presented that aims to take AI & music research a step further, going beyond surface features and focusing on the *expressive* aspects, and how these are communicated in music. We will look at recent work on computational models of expressive music performance and some examples of the state of the art, and will discuss possible applications of this research. In the process, the audience will be subjected to a little experiment which may, or may not, yield a surprising result.  View more, opens an external URL in a new window
 

Sep, 2019: CP’s researcher Christine Bauer on the Austrian public radio Ö1 on music and technology, and their impact on the music industry. Among other experts in the field, the Radiokolleg will feature CP’s researcher Christine Bauer in part 2 and 4 of the program, opens an external URL in a new window (1st and 3rd October 2019, respectively).
 

Sep, 2019: CP-JKU achieves first place in `emotion and theme recognition in music challenge, MediaEval 19. 
MediaEval is a benchmarking that offers challenges in multimedia retrieval, access and exploration to allow researchers working in computer science and other multimedia related field an opportunity to work on tasks that are related to human and social aspects of multimedia. MediaEval aims is to promote reproducible research that makes multimedia a positive force for society. 
Emotion and theme recognition is a popular task in music information retrieval that is relevant for music search and recommendation systems. The goal of the task is to automatically recognize the emotions and themes conveyed in a music recording using machine learning algorithms. See more here, opens an external URL in a new window

The team of the Institute of Computational Perception, namely Khaled Koutini, Shreyan Chowdhury, Verena Haunschmid, Hamid Eghbal-zadeh, Gerhard Widmer, won the challenge by developing the highest performing models to detect emotion and themes in music. The report detailing the work can be found here, opens a file
 

Sep, 2019: Best Reviewer Awards: CP researchers Christine Bauer and Markus Schedl receive two Best Reviewer Awards at ACM Recommender Systems Conference (out of more than 200 reviewers). A third Best Reviewer Award goes to former CP member Peter Knees.
 

Sep, 2019: CP as strong contributor to this year's Ars Electronica Festival: CP is a strong contributor to this year's *Ars Electronica Festival*, and the newly established *AI x Music* focus in particular.

"Computers that Learn to Listen, opens an external URL in a new window" - curated by Gerhard Widmer, opens in new window: a video that demonstrates some results of our research in the field of AI and Music, from autonomous drum robots to an automatic expressive accompaniment system.

"Recommenders and Intelligent Tools in Music Creation: Why, Why Not, and How?, opens an external URL in a new window" - a workshop by Christine Bauer, opens in new window, Peter Knees, Richard Vogl, opens in new window, and Hansi Raber - will highlight the role of AI, Machine Learning-supported composition, and Recommender Systems in the process of music creation.

"Music Information Retrieval: Inside and Outside the Music, opens an external URL in a new window" - a talk by Christine Bauer, opens in new window and Peter Knees: In scope of the AIxMusic Matinée, Christine Bauer and Peter Knees give a talk where they will highlight the use of intelligent audio analysis and social media mining for selected applications on music information retrieval and speculate on future directions.
 

Jul, 2019: Top Ranks for CP Teams in the DCASE 2019 Challenge: CP Teams achieve top ranks in the DCASE 2019 Challenge Task 1 and 2! 

CP Results in the DCASE 2019 Acoustic Event Detection and Classification Challenge, opens an external URL in a new window: 
Task 1a - Acoustic Scene Classification, opens an external URL in a new window.
Rank 2 (Team Ranking)  for CP Team Khaled Koutini, Hamid Eghbal-zadeh & Gerhard Widmer (see Report, opens an external URL in a new window)
Task 1b - Acoustic Scene Classification under Distribution Mismatch, opens an external URL in a new window:
Rank 3 for CP Team Hamid Eghbal-zadeh, Khaled Koutini, and Gerhard Widmer (Report, opens an external URL in a new window)
Rank 4 by CP Student! Team Paul Primus and David Eitelsebner (Report, opens an external URL in a new window)
Task 1c - Open-Set Acoustic Scene Classification, opens an external URL in a new window:
Rank 4 for SAL/CP Team Bernhard Lehner and Khaled Koutini (Report, opens an external URL in a new window)
Task 2 - Audio Tagging with Noisy Labels and Minimal Supervision, opens an external URL in a new window.
Rank 5 for CP Team Khaled Koutini, Hamid Eghbal-zadeh & Gerhard Widmer. (Report, opens an external URL in a new window)
 

Jun, 2019:  Study about Global and Country-specific Mainstreaminess Published: How mainstream are the listeners in my country and can we use this to improve music recommender systems? A recent study quantifies this "mainstreaminess" and presents music recommender systems which take this measure into account.

The researchers Christine Bauer and Markus Schedl published a study about global and country-specific mainstreaminess. In this, they analyze music preferences of 47 countries with respect to what is considered mainstream in a particular country. Furthermore, based on these findings, they evaluate recommendation approaches considering users’ mainstreaminess.

After publication, some of the main findings were summarized in these Austrian media articles (in German):
https://science.orf.at/stories/2987238/, opens an external URL in a new window

https://derstandard.at/2000105229715/Linzer-Forscher-zeigen-Laenderunterschiede-im-Musikgeschmack-auf, opens an external URL in a new window

https://futurezone.at/science/linzer-forscher-analysieren-musikgeschmack-in-verschieden-laendern/400529719, opens an external URL in a new window
For detailed information, please find the full article here, opens an external URL in a new window.
 

June, 2019: Con Espressione! Exhibit: The Con Espressione! Exhibit, opens an external URL in a new window is an interactive system designed for popular science exhibitions. It demonstrates and enables joint human-computer control of expressive performance: the visitor controls overall tempo and loudness of classical piano pieces (such as Beethoven's "Moonlight" sonata Op.27 No.2) via hand movements (tracked by a LeapMotion, opens an external URL in a new window sensor). In the background, our "Basis Mixer" expressive performance model adds subtle modifications to the performance, such as articulation and micro-timing (e.g., slight temporal differences in the note onsets when playing chords). The contribution of the Basis Mixer can be controlled and experimented with via a slider.

The exhibit was first shown at the la.la.lab Science Exhibition, opens an external URL in a new window ("The Mathematics of Music") in Heidelberg, Germany (May - Dec. 2019). The source code is openly available via a github repository, opens an external URL in a new window.
 

Okt, 2018: Best Paper and Best Poster Presentation Award:  Best paper award at the 19th International Society for Music Information Retrieval Conference.

The paper "Learning to Listen, Read, and Follow: Score Following as a Reinforcement Learning Game (PDF, opens an external URL in a new window)" by Matthias Dorfer, Florian Henkel, and Gerhard Widmer received the Best Paper Award and the Best Poster Presentation Award at the 19th International Society for Music Information Retrieval Conference, opens an external URL in a new window.
 

Sep, 2018: Best Paper Award: Best paper award at the 21st International Conference on Digital Audio Effects. The paper "Towards Multi-Instrument Drum Transcription" by Richard Vogl, opens an external URL in a new window (cp.jku PhD Student), Gerhard Widmer, opens in new window, and Peter Knees, opens an external URL in a new window (TU Wien) received the 2nd Best Paper Award at the 21st International Conference on Digital Audio Effects, opens an external URL in a new window. 
 

Sep, 2018: BE OPEN - Science & Society Festival: Gerhard Widmer to curate a science pavilion on "AI & Music" at the BE OPEN Science & Society Festival, opens an external URL in a new window,
Vienna, Sept. 8-12, 2018 (on the occasion of the 50th anniversary of the Austrian Science Fund (FWF), opens an external URL in a new window.
 

Sep, 2018: BE OPEN - Science & Society Festival, Article on our AI & Music Pavillion, opens an external URL in a new window at the FWF BE OPEN Science & Society Festival, opens an external URL in a new window produced by Austran Newspaper "Die Presse".
 

May, 2018: ERC = Science^2: On-line article in ERC = Science² ("Mind-blowing science, from the European Research Council") on our work on computational models of expressive music performance in the Con Espressione Project.
 

Mar, 2018: STANDARD Article: Article in DER STANDARD on the FWF project "Fine-grained culture-aware music recommender systems" led by cp.jku researcher Christine Bauer.
 

• Best Paper: "Bridging the Audio-Symbolic Gap: The Discovery of Repeated Note Content Directly From Polyphonic Music Audio, opens in new window" by T. Collins, S. Böck, F. Krebs, and G. Widmer
   
• Best Demo: "The Complete Classical Music Companion V0.9, opens in new window" by A. Arzt, S. Böck, S. Flossmann, H. Frostel, M. Gasser, and G. Widmer