AI Paper Atlas

Research map / Video / Action recognition

Progressive Cross-Modal Causal Intervention for Long-Term Action Recognition

Shaowu Xu, Xibin Jia, Chao Fan, Junyu Gao, Jing Chang, Qianmei Sun

CVPR 2026 · Video › Action recognition · approach: Disentangled action recognition

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 ICH (OT-based back-door intervention for co-occurrence hallucination) to baseline (ablation)PCMCI baseline · Breakfastaccuracy: 91.55 -> 94.93Improved
add ICI (relation-aware intervention) to baseline (ablation)PCMCI baseline · Breakfastaccuracy: 91.55 -> 93.80Improved
add IVC (front-door adjustment) to baseline (ablation)PCMCI baseline · BreakfastmAP: 80.23 -> 88.31Improved
add IVC on top of ICH and ICI (ablation)PCMCI · Breakfastaccuracy: 95.77 -> 97.46Improved
add IVC on top of ICH and ICI (ablation)PCMCI · COINmAP: 82.25 -> 86.54Improved
ICH alone versus ICI alone (ablation)PCMCI · COINaccuracy: 92.56 vs 91.74Improved
apply IVC first instead of last (ablation)PCMCI · Breakfastaccuracy: 97.46 -> 92.96Hurt
apply ICI before ICH (ablation)PCMCI · BreakfastmAP: 90.51 -> 87.43Hurt
use deconfounded text embeddings plus visual relational modeling as mediator instead of memory bank (ablation)PCMCI IVC · COINaccuracy: 90.60 -> 94.53Improved
add relational modeling in visual modality versus text-only mediator alignment (ablation)PCMCI IVC · BreakfastmAP: 88.75 -> 90.51Improved
use 8 relations in the relation-aware design (ablation)PCMCI · Breakfast and Charadesaccuracy and mAPImproved
replace CLIP with stronger BLIP2 backbone (ablation)PCMCI · COINaccuracy: 94.53 -> 94.89Improved

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