Mar 3, 2022 · We propose an end-to-end framework SegTAD composed of a 1D semantic segmentation network (1D-SSN) and a proposal detection network (PDN).

Feb 14, 2023 · We propose a unified framework SegTAD composed of a 1D semantic segmentation network (1D-SSN) and a proposal detection network (PDN).

SegTAD contains a 1D semantic segmentation network to learn the category of each single frame using the segmentation labels, which are directly transformed from ...

We propose an end-to-end temporal action detection framework to take advantage of the fine-grained prediction of semantic segmentation for more precise de-.

An end-to-end framework SegTAD is proposed composed of a 1D semantic segmentation network (1D-SSN) and a proposal detection network (PDN) that takes ...

Duration: 5:00
Posted: Oct 13, 2022

Segtad: Precise temporal action detection via semantic segmentation. Chen Zhao, Merey Ramazanova, Mengmeng Xu, Bernard Ghanem. January 2022.

(SegTAD）SegTAD: Precise Temporal Action Detection via Semantic Segmentation (arxiv 2022); (OpenTAL) OpenTAL: Towards Open Set Temporal Action Localization ...

Mar 3, 2022 · In this paper, we propose a novel perspective to for- mulate the task of temporal action detection based on 1D semantic segmentation to achieve ...

SegTAD: Precise Temporal Action Detection via Semantic Segmentation · Temporal action detection (TAD) is an important yet challenging task in video analysis.