C
Chris Pal
Researcher at École Polytechnique de Montréal
Publications - 262
Citations - 20801
Chris Pal is an academic researcher from École Polytechnique de Montréal. The author has contributed to research in topics: Computer science & Recurrent neural network. The author has an hindex of 57, co-authored 235 publications receiving 16589 citations. Previous affiliations of Chris Pal include University of Guelph & Association for Computing Machinery.
Papers
More filters
Posted Content
Fine-grained zero-shot recognition with metric rescaling.
TL;DR: It is demonstrated that the proposed instance-based deep metric learning approach, notwithstanding its simplicity of implementation and training, is superior to all the recent state-of-the-art methods of which the author is aware that use the same evaluation framework.
Proceedings Article
AR-DAE: Towards Unbiased Neural Entropy Gradient Estimation
TL;DR: The authors proposed the amortized residual denoising autoencoder (AR-DAE) to approximate the gradient of the log density function, which can be used to estimate the gradient gradient of entropy.
Posted Content
Medical Imaging with Deep Learning: MIDL 2020 - Short Paper Track.
TL;DR: This compendium gathers all the accepted extended abstracts from the Third International Conference on Medical Imaging with Deep Learning (MIDL 2020), held in Montreal, Canada, 6-9 July 2020.
Visual Imitation with Reinforcement Learning using Recurrent Siamese Networks
TL;DR: In this paper, a Siamese network is trained to compute distances between observed behaviors and an agent's behaviors, and then a comparator model is used to learn such distances in space and time between motion clips while training an RL policy to minimize this distance.
Journal ArticleDOI
A General Purpose Neural Architecture for Geospatial Systems
Nasim Rahaman,Martin Weiss,Frederik Trauble,Francesco Locatello,Aymeric Lacoste,Yoshua Bengio,Chris Pal,Li Erran Li,Bernhard Schölkopf +8 more
TL;DR: In this article , the authors present a roadmap towards the construction of a general-purpose neural architecture (GPNA) with a geospatial inductive bias, pre-trained on large amounts of unlabeled earth observation data in a self-supervised manner.