AI Paper Atlas

Research map / Video / Action recognition

Few-Shot Action Recognition with Hierarchical Matching and Contrastive Learning

Sipeng Zheng, Shizhe Chen, Qin Jin

ECCV 2022 · Video › Action recognition · approach: Disentangled action recognition · 73 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 cross-entropy fine-tuning with spatial-temporal transformer (ablation)pretrained ResNet-50 · Kinetics 1-shot / SSv2* 1-shotaccuracy: 59.9 -> 62.5; 29.5 -> 37.8Improved
train with global contrastive loss Lg alone instead of Lce (ablation)HCL · Kinetics 1-shotaccuracy: 62.5 -> 54.9Hurt
add global contrastive loss Lg to Lce (ablation)HCL · Kinetics 1-shotaccuracy: 62.5 -> 66.1Improved
add temporal contrastive loss with cycle consistency (Lt+Ltc) (ablation)HCL · SSv2* 1-shotaccuracy: 40.3 -> 45.4Improved
add spatial contrastive loss with cycle consistency (Ls+Lsc) (ablation)HCL · Kinetics 1-shotaccuracy: 66.1 -> 72.3Improved
add both temporal and spatial contrastive losses to Lce+Lg (ablation)HCL · Kinetics 1-shot / SSv2* 1-shotaccuracy: 66.1 -> 73.7; 40.3 -> 47.3Improved
add temporal cycle consistency loss Ltc (ablation)HCL · SSv2* 1-shot / 5-shotaccuracy: -1.5% for 1-shot and -0.9% for 5-shot without LtcImproved
add spatial cycle consistency loss Lsc (ablation)HCL · Kinetics 1-shotaccuracy: 72.1 -> 73.1Improved
add semantic attention in spatial matching (ablation)HCL · Kinetics 1-shot / 5-shotaccuracy: 71.3 -> 73.7; 84.4 -> 85.8Improved
add semantic attention in spatial matching (ablation)HCL · SSv2* 1-shotaccuracy: 46.3 -> 47.3No clear effect
add temporal and spatial zoom-in matching to global matching (cross-entropy training) (ablation)Lce-trained model · Kinetics 1-shotaccuracy: 62.5 -> 64.4Improved
add temporal and spatial zoom-in matching to global matching (HCL training) (ablation)HCL · Kinetics 1-shot / SSv2* 1-shotaccuracy: 69.1 -> 73.7; 42.9 -> 47.3Improved

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