AI Paper Atlas

Research map / Video / Video-language understanding

WeaveTime: Streaming from Earlier Frames into Emergent Memory in VideoLLMs

Yulin Zhang, Cheng Shi, Sibei Yang

CVPR 2026 · Video › Video-language understanding · approach: Long video understanding

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
finetune on offline video with timestamp prompts only (no temporal reconstruction) (ablation)LLaVA-OV-7B + ReKV · OVO-Benchoverall accuracy (%): 53.56 -> 49.88 (-3.68)Hurt
add Temporal Reconstruction auxiliary task (SOPE) to finetuning (ablation)LLaVA-OV-7B + ReKV · OVO-Benchoverall accuracy (%): 49.88 -> 55.70 (+5.82)Improved
add Temporal Reconstruction auxiliary task (SOPE) (ablation)LLaVA-OV-7B + ReKV · Streaming-Bench Real-Timeaccuracy (%): 65.91 -> 68.49 (+2.58)Improved
add PCDF-Cache (uncertainty-gated recall) on top of SOPE (ablation)LLaVA-OV-7B + ReKV · OVO-Benchoverall accuracy (%): 55.70 -> 57.57 (+1.87)Improved
add PCDF-Cache on top of SOPE (ablation)LLaVA-OV-7B + ReKV · Streaming-Bench Real-Timeaccuracy (%): 68.49 -> 72.13 (+3.64)Improved
use coarse-to-fine recall instead of coarse-only retrieval (ReKV) (ablation)LLaVA-OV · QAEgo4Drecall: 23.9 -> 25.2Improved
use coarse-to-fine recall instead of coarse-only retrieval (ablation)LLaVA-OV · MLVUaccuracy: 68.5 -> 68.9Improved
use fine-grained retrieval over full history instead of coarse-to-fine (ablation)LLaVA-OV · QAEgo4D/MLVU/EventHALLmemory: OOMHurt
raise entropy threshold for recall in PCDF-Cache (ablation)WeaveTime · OVO-Benchoverall accuracy / latency: accuracy peaks 57.57 at threshold 0.6; latency decreases monotonicallyMixed
add WeaveTime to LLaVA-OV-7B instead of ReKVLLaVA-OV-7B · OVO-Bench Real-Timeaverage accuracy (%): 61.72 -> 68.82Improved
use SOPE with 30K offline samples instead of stream-tailored dataLLaVA-OV + ReKV · OVO-Benchoverall accuracy (%): 53.6 -> 55.7 with 8 GPUs and 0 stream-tailored dataImproved

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