AI Paper Atlas

Research map / Video / Action recognition

Class Semantics-Based Attention for Action Detection

Deepak Sridhar, Niamul Quader, Srikanth Muralidharan, Yaoxin Li, Peng Dai, Juwei Lu

ICCV 2021 · Video › Action recognition · approach: Weakly supervised temporal action localization · 69 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add CSA attentionBMN · THUMOS-14avg mAP: 41.3 -> 47.5Improved
add CSA attentionGTAD · THUMOS-14avg mAP: 38.8 -> 42.6Improved
add CSA attentionBMN · ActivityNet v1.3avg mAP: 34.83 -> 35.43Improved
add CSA attentionBMN · ActivityNet v1.3 (high tIoU)mAP@tIoU0.95: 6.53 -> 5.18Hurt
add CSA attentionDBG · ActivityNet v1.3 proposalsAUC: 68.19 -> 68.89Improved
use CSA instead of SE attention at encoder output (ablation)BMN · THUMOS-14avg mAP: 44.2 -> 47.5Improved
use CSA instead of Transformer attention (ablation)BMN · THUMOS-14avg mAP: 39.1 -> 47.5Improved
add 1D convolution in attention instead of feed-forward (ablation)CSA · THUMOS-14avg mAP: 44.1 -> 46.7Improved
add temporal attention to channel attention (ablation)CSA · THUMOS-14avg mAP: 45.1 -> 46.7Improved
add channel attention to temporal attention (ablation)CSA · THUMOS-14avg mAP: 45.7 -> 46.7Improved
increase conv kernel size from 3 to 5 (ablation)CSA · THUMOS-14avg mAP: 46.7 -> 39.8Hurt
increase conv blocks from 1 to 2 (ablation)CSA · THUMOS-14avg mAP: 46.7 -> 47.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy