AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Starve to Perceive: Taming Lazy Perception in VLMs with Constrained Visual Bandwidth

Yuhuan Wu, Haozhe Wang, Cong Wei, Chong Peng, Fangzhen Lin, Wenhu Chen

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: VLM capability analysis · 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
train with constrained visual bandwidth (budget-aware SFT + RL) vs DeepEyesQwen2.5-VL-7B · Overall avg, unconstrainedoverall score: 61.77 -> 64.59Improved
train with constrained visual bandwidth vs DeepEyes, evaluated under B=256Qwen2.5-VL-7B · Overall avg, constrained B=256overall score: 45.57 -> 52.35Improved
use budget-aware SFT instead of standard SFT on identical trajectories (ablation)Qwen2.5-VL-7B SFT · HR-Bench 8Kaccuracy: 59.88 -> 65.88Improved
use budget-aware SFT instead of standard SFT (ablation)Qwen2.5-VL-7B SFT · V* Overallaccuracy: 75.92 -> 80.62Improved
use budget-aware SFT instead of standard SFT on V* relative position (ablation)Qwen2.5-VL-7B SFT · V* rela_posaccuracy: 73.68 -> 73.68No clear effect
skip budget-aware SFT warm start before budget RL (ZeroShot + BudgetRL) (ablation)Qwen2.5-VL-7B RL · HR visual search benchmarksaccuracy: lowest across all benchmarks (no numbers in text)Hurt
apply visual bandwidth constraint during RL instead of unconstrained RL (ablation)BA-SFT initialized RL · All Focus Ratioratio: 0.40 -> 0.50Improved
apply bandwidth constraint during RL (ablation)BA-SFT+RL · Unconstrained overall avgoverall score: 64.48 -> 64.59No clear effect
apply bandwidth constraint during RL (ablation)BA-SFT+RL · Constrained B=256 overall avgoverall score: 51.54 -> 52.35Improved
add RL on top of budget-aware SFT (ablation)BA-SFT · Unconstrained overall avgoverall score: 58.95 -> 64.59Improved
bandwidth-constrained RL vs vanilla RL across inference budgets 128-1024 (ablation)Ours · Budget sweep avgaverage accuracy: +1.38 (B=128), +0.81, +0.50, +0.60 (B=1024)Improved
constrained inference budget (B=256) vs unconstrainedOurs · Inference latencyspeedup: 2.7-5.1x speedupImproved

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