Entropy-driven Unsupervised Keypoint Representation Learning in Videos

TL;DR

This paper introduces an unsupervised method for learning meaningful video representations by leveraging local entropy to identify keypoints that capture dynamic information, improving object and scene understanding.

Contribution

The paper proposes a novel entropy-based unsupervised learning framework that discovers spatially and temporally consistent keypoints directly from video frames, guided by two information-theoretic losses.

Findings

Outperforms baselines in downstream tasks like object dynamics

Effectively captures static and dynamic objects in videos

Enhances understanding of scene changes and object movements

Abstract

Extracting informative representations from videos is fundamental for effectively learning various downstream tasks. We present a novel approach for unsupervised learning of meaningful representations from videos, leveraging the concept of image spatial entropy (ISE) that quantifies the per-pixel information in an image. We argue that \textit{local entropy} of pixel neighborhoods and their temporal evolution create valuable intrinsic supervisory signals for learning prominent features. Building on this idea, we abstract visual features into a concise representation of keypoints that act as dynamic information transmitters, and design a deep learning model that learns, purely unsupervised, spatially and temporally consistent representations \textit{directly} from video frames. Two original information-theoretic losses, computed from local entropy, guide our model to discover consistent keypoint representations; a loss that maximizes the spatial information covered by the keypoints and a loss that optimizes the keypoints' information transportation over time. We compare our keypoint representation to strong baselines for various downstream tasks, \eg, learning object dynamics. Our empirical results show superior performance for our information-driven keypoints that resolve challenges like attendance to static and dynamic objects or objects abruptly entering and leaving the scene.