AI Paper Atlas

Research map / Video / Video-language understanding

Bringing RNNs Back to Efficient Open-Ended Video Understanding

Weili Xu, Enxin Song, Wenhao Chai, Xuexiang Wen, Tian Ye, Gaoang Wang

ICCV 2025 · Video › Video-language understanding · approach: Long video understanding · 0 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
sort merged visual tokens in ascending size order instead of random order (ablation)AuroraLong · ActivityNet-QAaccuracy: 53.1 -> 56.3Improved
sort merged visual tokens in ascending order instead of descending (ablation)AuroraLong · ActivityNet-QAaccuracy: 55.0 -> 56.3Improved
sort merged visual tokens ascending instead of random (ablation)AuroraLong · VATEXBLEU@1: 67.6 -> 68.5Improved
sort merged visual tokens ascending instead of descending (ablation)AuroraLong · VATEXBLEU@1: 67.0 -> 68.5Improved
sort merged visual tokens ascending instead of random (ablation)AuroraLong · MovieChat-1Kaccuracy: 76.5 -> 78.5Improved
sort merged visual tokens ascending instead of descending (ablation)AuroraLong · MovieChat-1Kaccuracy: 76.0 -> 78.5Improved
sort merged visual tokens ascending instead of random (ablation)AuroraLong · VDC (average)accuracy: 40.9 -> 41.3Improved
sort merged visual tokens ascending instead of descending (ablation)AuroraLong · VDC (average)accuracy: 41.1 -> 41.3Improved
lower visual token retention ratio (more merging) for captioning tasks (ablation)AuroraLong · VATEX and VDCcaptioning scoreImproved
keep token retention at 0.1 instead of higher retention (ablation)AuroraLong · most video tasksaccuracyNo clear effect
Setting B (always merge tokens at ratio 0.1, including single images) instead of Setting A (merge only video samples) (ablation)AuroraLong · ANet, VATEX, MovieChat, VDCperformance: extra 50% training time for Setting AImproved
replace Transformer LLM with linear RNN (RWKV) backbone and token mergeAuroraLong vs InternVL-1.5 2B · 1024 framesGPU memory: 34X less memoryImproved

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