AI Paper Atlas

Research map / Video / Action recognition

DyFADet: Dynamic Feature Aggregation for Temporal Action Detection

Le Yang, Ziwei Zheng, Yizeng Han, Hao Cheng, Shiji Song, Gao Huang, Fan Li

ECCV 2024 · Video › Action recognition · approach: Disentangled action recognition · 35 citations (OpenAlex)

Read the paper

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
replace conv encoder with deformable conv (ablation)TAD baseline · THUMOS14avg mAP: 62.1 -> 66.1Improved
replace conv encoder with squeeze-and-excitation (ablation)TAD baseline · THUMOS14avg mAP: 62.1 -> 63.4Improved
replace conv with DFA_Conv in encoder (ablation)TAD baseline · THUMOS14avg mAP: 62.1 -> 66.8Improved
use DynE layer encoder instead of DFA_Conv (ablation)DyFADet · THUMOS14avg mAP: 66.8 -> 67.8Improved
add dynamic multi-scale head (DyHead) on top of DynE (ablation)DyFADet · THUMOS14avg mAP: 67.8 -> 69.2Improved
connect scales in head with plain convolution instead of DyHead (ablation)DyFADet · THUMOS14avg mAP: 67.8 -> 68.0No clear effect
add DyHead to a plain conv encoder (ablation)TAD baseline · THUMOS14avg mAP: 62.1 -> 65.8Improved
expansion factor w=5 vs 3 or 6 (ablation)DyFADet · THUMOS14avg mAP: w=3: 68.2, w=5: 69.2, w=6: 67.0Mixed
head depth D=3 vs 2 or 4 (ablation)DyFADet · THUMOS14avg mAP: D=2: 67.4, D=3: 69.2, D=4: 67.1Mixed
use C-formation vs K or CK formation for mask (ablation)DyFADet · THUMOS14avg mAP: 68.4 (C, CK) vs 69.2 (K)No clear effect
add optical flow features to VideoMAE V2-gDyFADet · THUMOS14avg mAP: 70.5 -> 71.1Improved
use DynE encoder without DyHead instead of self-attentionActionFormer · THUMOS14 latencyGPU latency (ms): 68.4 -> 44.7Improved

Later papers that compared with this method

3 later accepted papers report DyFADet as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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