In a join effort, Eric Guizzo and Tillman Weyde of MIRG and Francesco Barbieri, Giovanni Maffei, Frederico Luchesi and Fermin Moscoso del Prado Martin have developed a multi-modal emotion recognition system that won the OMG Empathy Challenge. See here and here for more information.
Read more →Tillman & Radha went to Montreal to attend NeurIPS 2018 and presented a Poster in the Relational Learning Workshop. Tillman also presented a Poster with Dan Philps and Artur Garcez in the Workshop on Financial Services and AI. NeurIPS had 8000 registered participants, a truly special experience.
Read more →Current and previous members of MIRG met at ISMIR 2018, presenting several papers. Former MIRG members are now working for Jukedeck, Spotify and Tido. From left to right: Shahar, Andreas (Spotify), Srikanth (Jukedeck), Daniel (Tido, City UoL), Reinier (Jukedeck), Tillman, Radha (City).
Read more →A belated note that our poster on “Singing Voice Separation with a Deep U-Net Convolutional Network” was selected for the Best Poster Award. Congrats to Andreas Jansson for the main work for designing a great poster (and for sticking an iPad on it for the demo).
Read more →Here are some audio examples to accompany our paper. All audio from the Free Music Archive. David Byrne – My Fair Lady Mix Download Separated instrumental Download Separated vocal Download Edith Frost Cars and Parties Mix Download Separated instrumental Download Separated vocal Download Emerald Park – At the Mall Mix Download Separated instrumental Download...
Read more →We have been successful, together with a great consortium consisting of Queen Mary UoL, CNRS Paris, University of Music Weimar, University of Illinois, Columbia University, in securing funding from the Digging Into Data Challenge for our new project Dig That Lick: Analysing Large-Scale Data For Melodic Patterns In Jazz Performances.
Read more →Tillman Weyde gave a talk entitled Towards Digital Musicology with Jazz Data and Metadata at Columbia University (New York) at the conference The Future of Jazz Discography.
Read more →Our paper Generating Time: Rhythmic Perception, Prediction and Production with Recurrent Neural Networks (by Andy Elmsley, Tillman Weyde and Newton Armstrong) has been published in the Journal of Creative Music Systems.
Read more →Our article The Digital Music Lab: A Big Data Infrastructure for Digital Musicology has appeared in the ACM Journal on Computing and Cultural Heritage’s Special Issue on Digital Infrastructure for Cultural Heritage See also here for the authors’ accepted version.
Read more →