CATER: A diagnostic dataset for Compositional Actions and TEmporal Reasoning

Abstract

Computer vision has undergone a dramatic revolution in performance, driven in large part through deep features trained on large-scale supervised datasets. However, much of these improvements have focused on static image analysis; video understanding has seen rather modest improvements. Even though new datasets and spatiotemporal models have been proposed, simple frame-by-frame classification methods often still remain competitive. We posit that current video datasets are plagued with implicit biases over scene and object structure that can dwarf variations in temporal structure. In this work, we build a video dataset with fully observable and controllable object and scene bias, and which truly requires spatiotemporal understanding in order to be solved. Our dataset, named CATER, is rendered synthetically using a library of standard 3D objects, and tests the ability to recognize compositions of object movements that require long-term reasoning. In addition to being a challenging dataset, CATER also provides a plethora of diagnostic tools to analyze modern spatiotemporal video architectures by being completely observable and controllable. Using CATER, we provide insights into some of the most recent state of the art deep video architectures.

arXiv commentICLR 2020 (oral). Webpage/code/data: https://rohitgirdhar.github.io/CATER

References in corpus (11)

• Two-Stream Convolutional Networks for Action Recognition in Videos
• Playing Atari with Deep Reinforcement Learning
• UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild
• The Kinetics Human Action Video Dataset
• Unsupervised Learning of Video Representations using LSTMs
• Learning from Synthetic Humans
• Spatiotemporal Residual Networks for Video Action Recognition
• AI2-THOR: An Interactive 3D Environment for Visual AI
• Learnable pooling with Context Gating for video classification
• Revisiting the Effectiveness of Off-the-shelf Temporal Modeling Approaches for Large-scale Video Classification
• IntPhys: A Framework and Benchmark for Visual Intuitive Physics Reasoning

Cited by in corpus (8)

• CLEVRER: CoLlision Events for Video REpresentation and Reasoning
• LanguageRefer: Spatial-Language Model for 3D Visual Grounding
• Enhancing Unsupervised Video Representation Learning by Decoupling the Scene and the Motion
• DramaQA: Character-Centered Video Story Understanding with Hierarchical QA
• SORNet: Spatial Object-Centric Representations for Sequential Manipulation
• SPACE: A Simulator for Physical Interactions and Causal Learning in 3D Environments
• Learning Object Permanence from Video
• Maintaining Common Ground in Dynamic Environments