The 2015 ICISP Conference
16 Feb 2016
The conference ICISP was held in Trois-Rivières, Québec, Canada. The aim of ICISP 2016 is to report on the latest developments in image and signal processing, multimedia and computer graphics.
During the three day conference, three keynote speeches were presented. Professor Robert Laganiere gave a talk titled “Smart surveillance systems: trends and approached”. This talk mainly discussed some of the challenges related to the development for the next generation of video surveillance systems. Professor Jocelyn Channusot gave a talk about the hierarchical analysis of multimodal data. In this talk, the techniques for extending the hierarchical representations to multimodal images in order to better exploit the additional information brought by the multimodality and improve classical image techniques are investigated. Dr. Roland Memisevic presented the use of deep learning techniques for solving image processing problems. Both the role of deep learning and main challenges are presented. Meanwhile, recent publications in the domain of applying deep learning for signal and image processing and some ideals and approaches to address the aforementioned challenges were discussed.
I presented two papers at this conference, which are titled Multiple-Instance Multiple-Label Learning for the Classification of Frog Calls with Acoustic Event Detection (link) and Feature Extraction Based on Bandpass Filtering for Frog Call Classification (link). The people at the conference were very interested in my work, asked several questions, and offered feedback for improving my current work. After the conference, I spent one day exploring the city of Trois-Rivières. I really enjoyed the French culture in Québec.