We held our tinyML Talks webcast with one presentation: Vincent Gripon has presented A Review of Compression Methods for Deep Convolutional Neural Network on July 17, 2020 at 8:00 AM Pacific Time.
We are very excited to present the first tinyML Montreal talk with a fantastic speaker and topic! In this talk, Vincent will share his expertise and research as we review the main methods to compress and accelerate Convolutional Neural Networks (CNNs): pruning, distillation, clustering, quantization, etc. to deploy CNNs and make them applicable for tinyML devices. We will present both pros and cons of each technique in an attempt to better understand how to reduce the size of such architectures to run on device without compromising the accuracy. There will also be a Q&A session at the end of the talk for the audience to ask questions.
Vincent Gripon is a Permanent Researcher at IMT-Atlantique, an Invited Professor at Université Côte d’Azur and was a Visiting Researcher at MILA from Oct. 2018 to Sept. 2019.
Watch on YouTube:
Download presentation slide:
Feel free to ask your questions on this thread and keep the conversation going!