Made to Order: Discovering monotonic temporal changes via self-supervised video ordering

TL;DR

This paper introduces a self-supervised transformer model that learns to identify and localize monotonic temporal changes in image sequences by leveraging sequence ordering as a proxy task, with applications across various domains.

Contribution

The paper presents a novel transformer-based approach for discovering monotonic temporal changes using self-supervised sequence ordering, with built-in attribution maps for localization.

Findings

Successfully discovers and localizes monotonic changes in diverse sequences

Attention maps serve as effective prompts for segmenting changing regions

Achieves state-of-the-art results on image ordering benchmarks

Abstract

Our objective is to discover and localize monotonic temporal changes in a sequence of images. To achieve this, we exploit a simple proxy task of ordering a shuffled image sequence, with `time' serving as a supervisory signal, since only changes that are monotonic with time can give rise to the correct ordering. We also introduce a transformer-based model for ordering of image sequences of arbitrary length with built-in attribution maps. After training, the model successfully discovers and localizes monotonic changes while ignoring cyclic and stochastic ones. We demonstrate applications of the model in multiple domains covering different scene and object types, discovering both object-level and environmental changes in unseen sequences. We also demonstrate that the attention-based attribution maps function as effective prompts for segmenting the changing regions, and that the learned representations can be used for downstream applications. Finally, we show that the model achieves the state-of-the-art on standard benchmarks for image ordering.