output
20112026
most citedGeometric deep learning: going beyond Euclidean data

3.7k citations

Showing 2017Show all

8 papers · 1 filter

cs.CL2017★ 34 cited

Letter-Based Speech Recognition with Gated ConvNets

Vitaliy Liptchinsky, Gabriel Synnaeve, Ronan Collobert

In the recent literature, "end-to-end" speech systems often refer to letter-based acoustic models trained in a sequence-to-sequence manner, either via a recurrent model or via a st…

cs.CV2017★ 28 cited

Detect-and-Track: Efficient Pose Estimation in Videos

Rohit Girdhar, Georgia Gkioxari, Lorenzo Torresani +2

This paper addresses the problem of estimating and tracking human body keypoints in complex, multi-person video. We propose an extremely lightweight yet highly effective approach t…

cs.CV2017★ 13 cited

CoDraw: Collaborative Drawing as a Testbed for Grounded Goal-driven Communication

Jin-Hwa Kim, Nikita Kitaev, Xinlei Chen +5

In this work, we propose a goal-driven collaborative task that combines language, perception, and action. Specifically, we develop a Collaborative image-Drawing game between two ag…

cs.LG2017★ 8 cited

A hierarchical loss and its problems when classifying non-hierarchically

Cinna Wu, Mark Tygert, Yann LeCun

Failing to distinguish between a sheepdog and a skyscraper should be worse and penalized more than failing to distinguish between a sheepdog and a poodle; after all, sheepdogs and…

cs.LO2017

Proof-Relevant Logical Relations for Name Generation

Nick Benton, Martin Hofmann, Vivek Nigam

Pitts and Stark's -calculus is a paradigmatic total language for studying the problem of contextual equivalence in higher-order languages with name generation. Models for the $ν…

cs.GT2017★ 4 cited

Demand-Independent Optimal Tolls

Riccardo Colini-Baldeschi, Max Klimm, Marco Scarsini

Wardrop equilibria in nonatomic congestion games are in general inefficient as they do not induce an optimal flow that minimizes the total travel time. Network tolls are a prominen…