Long-Tail Temporal Action Segmentation with Group-wise Temporal Logit Adjustment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pang, Zhanzhong, Sener, Fadime, Ramasubramanian, Shrinivas, Yao, Angela
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914915855368192
author Pang, Zhanzhong
Sener, Fadime
Ramasubramanian, Shrinivas
Yao, Angela
author_facet Pang, Zhanzhong
Sener, Fadime
Ramasubramanian, Shrinivas
Yao, Angela
contents Procedural activity videos often exhibit a long-tailed action distribution due to varying action frequencies and durations. However, state-of-the-art temporal action segmentation methods overlook the long tail and fail to recognize tail actions. Existing long-tail methods make class-independent assumptions and struggle to identify tail classes when applied to temporal segmentation frameworks. This work proposes a novel group-wise temporal logit adjustment~(G-TLA) framework that combines a group-wise softmax formulation while leveraging activity information and action ordering for logit adjustment. The proposed framework significantly improves in segmenting tail actions without any performance loss on head actions.
format Preprint
id arxiv_https___arxiv_org_abs_2408_09919
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Long-Tail Temporal Action Segmentation with Group-wise Temporal Logit Adjustment
Pang, Zhanzhong
Sener, Fadime
Ramasubramanian, Shrinivas
Yao, Angela
Computer Vision and Pattern Recognition
Procedural activity videos often exhibit a long-tailed action distribution due to varying action frequencies and durations. However, state-of-the-art temporal action segmentation methods overlook the long tail and fail to recognize tail actions. Existing long-tail methods make class-independent assumptions and struggle to identify tail classes when applied to temporal segmentation frameworks. This work proposes a novel group-wise temporal logit adjustment~(G-TLA) framework that combines a group-wise softmax formulation while leveraging activity information and action ordering for logit adjustment. The proposed framework significantly improves in segmenting tail actions without any performance loss on head actions.
title Long-Tail Temporal Action Segmentation with Group-wise Temporal Logit Adjustment
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.09919