REVIEW 13 cited by
Divide, Conquer and Combine: A Training-Free Framework for High-Resolution Image Perception in Multimodal Large Language Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Divide, Conquer and Combine: A Training-Free Framework for High-Resolution Image Perception in Multimodal Large Language Models
read the original abstract
Multimodal large language models (MLLMs) have experienced significant advancements recently, but still struggle to recognize and interpret intricate details in high-resolution (HR) images effectively. While state-of-the-art (SOTA) MLLMs claim to process images at 4K resolution, existing MLLM benchmarks only support up to 2K, leaving the capabilities of SOTA models on true HR images largely untested. Furthermore, existing methods for enhancing HR image perception in MLLMs rely on computationally expensive visual instruction tuning. To address these limitations, we introduce HR-Bench, the first deliberately designed benchmark to rigorously evaluate MLLM performance on 4K&8K images. Through extensive experiments, we demonstrate that while downsampling HR images leads to vision information loss, leveraging complementary modalities, e.g., text, can effectively compensate for this loss. Building upon this insight, we propose Divide, Conquer and Combine (DC$^2$), a novel training-free framework for enhancing MLLM perception of HR images. DC$^2$ follows a three-staged approach: 1) Divide: recursively partitioning the HR image into patches and merging similar patches to minimize computational overhead, 2) Conquer: leveraging the MLLM to generate accurate textual descriptions for each image patch, and 3) Combine: utilizing the generated text descriptions to enhance the MLLM's understanding of the overall HR image. Extensive experiments show that: 1) the SOTA MLLM achieves 63% accuracy, which is markedly lower than the 87% accuracy achieved by humans on HR-Bench; 2) our DC$^2$ brings consistent and significant improvements (a relative increase of +6% on HR-Bench and +8% on general multimodal benchmarks). The benchmark and code will be released to facilitate the multimodal R&D community.
Forward citations
Cited by 13 Pith papers
-
UltraVR: A Diagnostic Ultra-Resolution Image-VQA Benchmark for Evidence-Grounded Reasoning
UltraVR is a new diagnostic benchmark for evidence-grounded VQA on ultra-resolution images, with structured chain-of-thought annotations that localize failures in grounding, perception, and inference.
-
VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation
Counterfactual present/removed teacher views attribute visually supported corrections and reconstruct student-anchored distillation targets that beat source-mixed multimodal OPD.
-
RP-OPSD: Resolution-Privileged On-Policy Self-Distillation for Multimodal Large Language Models
Using original-resolution images as privileged teacher context over half-resolution student rollouts improves Qwen3.5 MLLMs by ~5.5% relative average score and trains 1.78× faster than answer-hint OPSD.
-
AdaTurn: Budget-Aware Test-Time Scaling for Active Visual Perception Agents
Budget-conditioned forced-answer RL lifts 4-turn VisualProbe-Medium from 36.7% to 47.6% while keeping 32-turn performance competitive.
-
ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison
ClaimDiff-RL replaces holistic scalar rewards with reference-conditioned atomic claim differences verified by a multimodal judge to improve the hallucination-missing-fact tradeoff in long-form image captioning.
-
ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison
ClaimDiff-RL introduces reference-conditioned atomic claim differences verified by a multimodal judge as the reward signal for fine-grained RL in long-form image captioning.
-
Starve to Perceive: Taming Lazy Perception in VLMs with Constrained Visual Bandwidth
Constraining visual token budgets during SFT and RL forces VLMs to learn functional active perception, yielding ~5% relative gains and strong transfer to unconstrained evaluation.
-
MCMit: Mid-Circuit Measurement Error Mitigation
MCMit proposes a constant-latency multi-control branch instruction, transformer and CNN discriminators, plus static MCM elimination and stochastic branching, evaluated on Qubic with QPU traces to cut latency by 70% an...
-
SIEVES: Selective Prediction Generalizes through Visual Evidence Scoring
SIEVES improves selective prediction coverage by up to 3x on OOD VQA benchmarks by training a selector to score the quality of visual evidence produced by reasoner models, generalizing across benchmarks and proprietar...
-
SIEVES: Selective Prediction Generalizes through Visual Evidence Scoring
SIEVES improves selective prediction coverage up to 3x on OOD VQA benchmarks by training a selector on visual localization quality, generalizing across datasets and proprietary reasoners without specific adaptation.
-
Adaptive Chain-of-Focus Reasoning via Dynamic Visual Search and Zooming for Efficient VLMs
Chain-of-Focus enables VLMs to adaptively search and zoom on important image areas via a two-stage SFT and RL pipeline on a custom 3K-sample dataset, yielding 5% gains on the V* benchmark across resolutions from 224 to 4K.
-
Starve to Perceive: Taming Lazy Perception in VLMs with Constrained Visual Bandwidth
Constraining visual token budget per observation during VLM training forces genuine active perception and delivers 5% average relative improvement without auxiliary losses or architecture changes.
-
MCMit: Mid-Circuit Measurement Error Mitigation
MCMit mitigates mid-circuit measurement errors via a new multi-control branch instruction, CNN and transformer discriminators, and software techniques, reporting up to 70% latency reduction and 80% lower logical error...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.