ERMITES 2016 is the first int. Bioacoutic Hackathon : you'll have to solve in team some of the most challenging classification and tracking paradigm in soundscape analysis : bioacoustic transient analysis. The material you'll work on is from endangered species front of Toulon : Physter Macrocephalus, recorded by several observatories in our project VAMOS. This international leading event brings researchers, students, ingeneers, and provides participants the opportunity to gain deeper insight into advanced research trends in scaled acoustics information retrieval within an interdisciplinary framework. It is organized as a series of 2 hours session, with long talks from experts (G. Pavan Cibra, H. Glotin LSIS, ... ), and solving sessions with brainstorming, massive computation and then debriffing during which attendees are invited to interact. (links to online videos of previous editions).
The goal of this edition is to learn and train yourself with expert worksing with you 2 full days to improve the performance of
soundscape or bioacoustic pattern detection and classification, at low
signal to noise ratio tansients.
You'll learn (a) how to learn bioacoustic
representation in the Deep Learning framework
(b) to develop more efficient supervised or unsupervised classification
of complex bioacoustic patterns.
Extensive computation solution and scalable solutions will be awarded.
At the end of the workshop, benchmark will be run to rank the challengers solutions.
Keywords: Learning Representation, Machine learning, Estimation, Unsupervised learning, Big data science, Signal processing, Wavelets, Deep learning, Sequence analysis, Statistics, Software platforms, Python, GPU programming, Passive acoustic hardware ...
ERMITES is supported by MADICS CNRS, Toulon university, SABIOD MASTODONS CNRS project, IUF, LSIS