AI Paper Atlas

Research map / Video / Action recognition

Self-supervised Motion Representation via Scattering Local Motion Cues

Yuan Tian, Zhaohui Che, Wenbo Bao, Guangtao Zhai, Zhiyong Gao

ECCV 2020 · Video › Action recognition · approach: Contrastive video representation learning · 21 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
replace nearest upsampling with learnable CGMUL (ablation)motion map upsampling · UCF-Flowl1 error (lower better): 0.042 -> 0.024Improved
replace bilinear upsampling with learnable CGMUL (ablation)motion map upsampling · UCF-Flowl1 error (lower better): 0.037 -> 0.024Improved
add CGMUL-upsampled HR motion map (vs bilinear) on top of LR (ablation)3D-ResNet18 · UCF101top-1 accuracy: 88.2% -> 90.6%Improved
add nearest or bilinear upsampled HR motion map on top of LR (ablation)3D-ResNet18 · UCF101top-1 accuracy: 88.2% -> 88.2%No clear effect
add LR motion map to the backbone (ablation)3D-ResNet18 · UCF101top-1 accuracy: 84.4% -> 88.2%Improved
train backbone with motion representation learning (no extra inference cost) (ablation)3D-ResNet18 · UCF101top-1 accuracy: 84.4 -> 86.5Improved
add HR motion pipeline fused with backbone (ablation)3D-ResNet18 · UCF101top-1 accuracy: 88.2 -> 89.4Improved
add LR+HR fusion instead of HR alone (ablation)3D-ResNet18 · UCF101top-1 accuracy: 89.4 -> 90.6Improved
use shallower 3D-ResNet18 with LR+HR motion rather than 3D-ResNet34 baselineaction recognition · UCF101top-1 accuracy: 87.7 -> 90.6Improved
increase motion-map upsampling scale factor to 4 from 1 (ablation)CGM-Net · action recognition (Fig. 6)top-1 accuracy: 6.2%Improved
increase upsampling scale factor beyond 8 (e.g. 16) (ablation)CGM-Net · motion estimation / action recognition (Fig. 6)l1 error / top-1Hurt
add CGM-Net multi-task motion learning on 3D-ResNeXt101 vs DynamoNet3D-ResNeXt101 · Kinetics-400top-1 accuracy: 68.2 -> 76.2Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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