Saved in:
Bibliographic Details
Main Authors: Fish, Edward, Gilbert, Andrew
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.18915
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912498539560960
author Fish, Edward
Gilbert, Andrew
author_facet Fish, Edward
Gilbert, Andrew
contents Few-shot temporal action localization (TAL) methods that adapt large models via single-prompt tuning often fail to produce precise temporal boundaries. This stems from the model learning a non-discriminative mean representation of an action from sparse data, which compromises generalization. We address this by proposing a new paradigm based on multi-prompt ensembles, where a set of diverse, learnable prompts for each action is encouraged to specialize on compositional sub-events. To enforce this specialization, we introduce PLOT-TAL, a framework that leverages Optimal Transport (OT) to find a globally optimal alignment between the prompt ensemble and the video's temporal features. Our method establishes a new state-of-the-art on the challenging few-shot benchmarks of THUMOS'14 and EPIC-Kitchens, without requiring complex meta-learning. The significant performance gains, particularly at high IoU thresholds, validate our hypothesis and demonstrate the superiority of learning distributed, compositional representations for precise temporal localization.
format Preprint
id arxiv_https___arxiv_org_abs_2403_18915
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PLOT-TAL: Prompt Learning with Optimal Transport for Few-Shot Temporal Action Localization
Fish, Edward
Gilbert, Andrew
Computer Vision and Pattern Recognition
Machine Learning
Few-shot temporal action localization (TAL) methods that adapt large models via single-prompt tuning often fail to produce precise temporal boundaries. This stems from the model learning a non-discriminative mean representation of an action from sparse data, which compromises generalization. We address this by proposing a new paradigm based on multi-prompt ensembles, where a set of diverse, learnable prompts for each action is encouraged to specialize on compositional sub-events. To enforce this specialization, we introduce PLOT-TAL, a framework that leverages Optimal Transport (OT) to find a globally optimal alignment between the prompt ensemble and the video's temporal features. Our method establishes a new state-of-the-art on the challenging few-shot benchmarks of THUMOS'14 and EPIC-Kitchens, without requiring complex meta-learning. The significant performance gains, particularly at high IoU thresholds, validate our hypothesis and demonstrate the superiority of learning distributed, compositional representations for precise temporal localization.
title PLOT-TAL: Prompt Learning with Optimal Transport for Few-Shot Temporal Action Localization
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2403.18915