AI Paper Atlas

Research map / Video / Video-language understanding

UTOPI: Efficient Egocentric Long-Video Understanding in AR via User-Guided Token Pre-Compression

ziqi wang, Su Chen, Qiance Tang, Jieyu Lin, Ziyun Li, Barbara De Salvo, Sai Qian Zhang

NeurIPS 2026 · Video › Video-language understanding · approach: Egocentric video understanding · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes UTOPI, a plug-and-play user-guided token pre-compression module that reduces redundant visual tokens in egocentric long videos for resource-limited AR devices.

Problem

Egocentric long-video understanding on AR devices is constrained by heavy memory and compute demands caused by substantial temporal redundancy in first-person videos.

Approach

A plug-and-play pre-compression module that uses user guidance (e.g., the wearer's attention/interaction cues) to pre-compress visual tokens before they reach the video model, tailored to egocentric streams.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
apply UTOPI with 15% token budgetQwen2.5VL-7B · EgoEye, EgoSchema, EgoGazeVQAaccuracy (%): EE 58.4 -> 60.6; ES 60.1 -> 64.8; EGV 47.8 -> 49.0Improved
use random token drop at 15% budgetQwen2.5VL-7B · EgoEye, EgoSchema, EgoGazeVQAaccuracy (%): EE 58.4 -> 56.8; ES 60.1 -> 59.3; EGV 47.8 -> 46.1Hurt
use frame subsampling at 15% budgetQwen2.5VL-7B · EgoEyeaccuracy (%): 58.4 -> 48.8Hurt
use UTOPI instead of DTD/PruneVid at 15% budgettoken compression · EgoEyeaccuracy (%): 57.6/51.3 -> 60.6Improved
add UTOPI to Qwen2.5VL-7B at extreme budget 3%Qwen2.5VL-7B · EgoEyeaccuracy (%): 58.4 -> 50.5Hurt
add UTOPI at 5% budgetQwen2.5VL-7B · EgoSchemaaccuracy (%): 60.1 -> 61.4Improved
add UTOPI to PLLaVA at 15% budgetPLLaVA · EgoEyeaccuracy (%): 34.8 -> 33.5Hurt
enable AGTS alone (ablation)UTOPI · EgoSchemaaccuracy (%): 60.1 -> 64.1Improved
enable IVOC alone (ablation)UTOPI · EgoSchemaaccuracy (%): 60.1 -> 64.4Improved
enable both AGTS and IVOC (ablation)UTOPI · EgoSchemaaccuracy (%) and token budget: 64.4 -> 64.8 at 15% budgetImproved
inject gaze noise up to 300 px (ablation)UTOPI · EgoEyeaccuracy (%): 60.6 -> 57.7Hurt
replace ground-truth gaze with SUM-predicted gaze (ablation)UTOPI · EgoEyeaccuracy (%): 60.6 -> 59.4No clear effect

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