Seminar of the Machine Listening, Machine Vision and Models of Sensory Neuroscience
Organizers: Jörg Lücke, Jörn Anemüller
Time and place:
Thursday: 12:00 - 14:00, Room: W32 1-113
Building up on advanced Machine Learning knowledge, this seminar discusses recent scientific contributions and developments in Machine Learning as well as recent papers on applications of Machine Learning algorithms. Typical application domains include general pattern recognition, computer hearing, computer vision and computational neuroscience. Typical tasks include auditory and visual signal enhancements, source separation, auditory and visual object learning and recognition, auditory scene analysis, data compression and inpainting. Applications to computationalneuroscience will discuss recent papers on the probabilistic interpretation of neural learning and biological intelligence.
|May 7th||Dictionary Learning with Occlusion and Masks||Jörg Lücke|
|May 21st||Conference Recaps (ICASSP 2015 and COSYNE 2015)||Saboor, Raphael, ICASSP attendees|
|May 28th||Approaches to Source Localization (acoustic, EEG, MEG)||Hendrik and Cris|
|June 4th||Collision with hearing aid developers forum|
|June 11th||Deep classification on acoustic data/ECog data||Bernd and Marina|
|June 18th||T1: Human-level control through deep reinforcement learning, T2: Deep Learning for Acoustic Modelling||Annika and Constantin (maybe Dennis)|
|July 2nd||Sparse coding, NMF, dictionary learning, codebook learning, probabilistic approaches||Kamil, Jörg, Maryam, Georgios|
|Signal enhancement and preprocessing||Timo, Kamil|
The students will learn the current research directions and challenges of the Machine Learning research field. By presenting examples from Machine Learning algorithms applied to sensory data tasks including task in Computer Hearing and Computer Vision the students will be taught the current strengths and weaknesses of different approaches. The presentations of current research papers by the participants will make use of computers and projectors. Programming examples and animations will be used to support the interactive component of the presentations. In scientific discussions of the presented and related work, the students will deepen their knowledge aboutcurrent limitations of Machine Learning approaches both on the theoretical side and on the side of their technical and practical realizations. Presentations of interdisciplinary research will enable the students to carry over their Machine Learning knowledge to address questions in other scientific domains.
- Standard ML Journals: JMLR, TPAMI, Neural Comp etc. (Will be distributed during the course.)