Semantically Guided Representation Learning For Action Anticipation
It addresses the problem of forecasting future activities in video sequences for applications like robotics and surveillance, with a focus on semantic interconnectivity, though it builds incrementally on prior representation learning methods.
The paper tackles action anticipation by learning semantically aware action representations using visual prototypes and language models, achieving improvements of +3.5, +2.7, and +3.5 points in Top-1 Accuracy on three benchmarks and +0.8 in Top-5 Recall on another.
Action anticipation is the task of forecasting future activity from a partially observed sequence of events. However, this task is exposed to intrinsic future uncertainty and the difficulty of reasoning upon interconnected actions. Unlike previous works that focus on extrapolating better visual and temporal information, we concentrate on learning action representations that are aware of their semantic interconnectivity based on prototypical action patterns and contextual co-occurrences. To this end, we propose the novel Semantically Guided Representation Learning (S-GEAR) framework. S-GEAR learns visual action prototypes and leverages language models to structure their relationship, inducing semanticity. To gather insights on S-GEAR's effectiveness, we test it on four action anticipation benchmarks, obtaining improved results compared to previous works: +3.5, +2.7, and +3.5 absolute points on Top-1 Accuracy on Epic-Kitchen 55, EGTEA Gaze+ and 50 Salads, respectively, and +0.8 on Top-5 Recall on Epic-Kitchens 100. We further observe that S-GEAR effectively transfers the geometric associations between actions from language to visual prototypes. Finally, S-GEAR opens new research frontiers in anticipation tasks by demonstrating the intricate impact of action semantic interconnectivity.