HERO: Rethinking Visual Token Early Dropping in High-Resolution Large Vision-Language Models

• URL: http://arxiv.org/abs/2509.13067v2
• Date: Mon, 29 Sep 2025 07:44:42 GMT
• Title: HERO: Rethinking Visual Token Early Dropping in High-Resolution Large Vision-Language Models
• Authors: Xu Li, Yuxuan Liang, Xiaolei Chen, Yi Zheng, Haotian Chen, Bin Li, Xiangyang Xue,
• Abstract summary: We propose HERO, a framework that integrates content-adaptive token budget allocation with function-aware token selection.<n>This study provides both empirical insights and practical solutions toward efficient inference in HR-LVLMs.
• Score: 60.028070589466445
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: By cropping high-resolution images into local tiles and encoding them independently, High-Resolution Large Vision-Language Models (HR-LVLMs) have demonstrated remarkable fine-grained visual understanding capabilities. However, this divide-and-conquer paradigm significantly increases the number of visual tokens, resulting in substantial computational and memory overhead. To better understand and address this challenge, we empirically investigate visual token utilization in HR-LVLMs and uncover three key findings: (1) the local tiles have varying importance, jointly determined by visual saliency and task relevance; (2) the CLS token in CLIP-based vision encoders exhibits a two-stage attention pattern across layers, with each stage attending to different types of visual tokens; (3) the visual tokens emphasized at different stages encode information at varying levels of granularity, playing complementary roles within LVLMs. Building on these insights, we propose HERO, a High-resolution visual token early dropping framework that integrates content-adaptive token budget allocation with function-aware token selection. By accurately estimating tile-level importance and selectively retaining visual tokens with complementary roles, HERO achieves superior efficiency-accuracy trade-offs across diverse benchmarks and model scales, all in a training-free manner. This study provides both empirical insights and practical solutions toward efficient inference in HR-LVLMs.

Related papers

• ViTCoP: Accelerating Large Vision-Language Models via Visual and Textual Semantic Collaborative Pruning [8.933549837045932]
  Large Vision-Language Models incur high computational costs due to significant redundancy in their visual tokens.<n>We propose a Visual and Textual Semantic Collaborative Pruning framework (ViTCoP) that combines redundancy filtering in the vision encoder with step-wise co-pruning within the Large Language Models.
  arXiv  Detail & Related papers  (2026-01-25T12:47:30Z)
• Attention Guided Alignment in Efficient Vision-Language Models [56.20286899428444]
  Large Vision-Language Models (VLMs) rely on effective multimodal alignment between pre-trained vision encoders and Large Language Models (LLMs)<n>This paper presents a comprehensive analysis of attention patterns in efficient VLMs.<n>We introduce Attention-Guided Efficient Vision-Language Models (AGE-VLM), a novel framework that enhances visual grounding through interleaved cross-attention layers.
  arXiv  Detail & Related papers  (2025-11-21T21:36:48Z)
• SECOND: Mitigating Perceptual Hallucination in Vision-Language Models via Selective and Contrastive Decoding [5.976839106353883]
  SECOND: Selective and Contrastive Decoding is a novel approach that enables Vision-Language Models to leverage multi-scale visual information with an object-centric manner.<n> SECOND significantly reduces perceptual hallucinations and outperforms a wide range of benchmarks.
  arXiv  Detail & Related papers  (2025-06-10T02:55:38Z)
• ToDRE: Visual Token Pruning via Diversity and Task Awareness for Efficient Large Vision-Language Models [59.47738955960352]
  ToDRE is a two-stage and training-free token compression framework.<n>It achieves superior performance by pruning tokens based on token Diversity and token-task RElevance.
  arXiv  Detail & Related papers  (2025-05-24T15:47:49Z)
• Instruction-Guided Fusion of Multi-Layer Visual Features in Large Vision-Language Models [50.98559225639266]
  We investigate the contributions of visual features from different encoder layers using 18 benchmarks spanning 6 task categories.<n>Our findings reveal that multilayer features provide complementary strengths with varying task dependencies, and uniform fusion leads to suboptimal performance.<n>We propose the instruction-guided vision aggregator, a module that dynamically integrates multi-layer visual features based on textual instructions.
  arXiv  Detail & Related papers  (2024-12-26T05:41:31Z)
• [CLS] Token Tells Everything Needed for Training-free Efficient MLLMs [66.5266435598799]
  Multi-language Large Language Models (MLLMs) have recently demonstrated strong performance across a wide range of vision tasks.<n>However, their efficient deployment remains a substantial challenge due to high computational costs and memory requirements.<n>We introduce a simple yet effective method for train-free visual compression, called VTC- compression.
  arXiv  Detail & Related papers  (2024-12-08T05:29:39Z)
• FoPru: Focal Pruning for Efficient Large Vision-Language Models [11.36025001578531]
  We propose Focal Pruning (FoPru), a training-free method that prunes visual tokens based on the attention-based token significance derived from the vision encoder. Our method can prune a large number of redundant tokens while maintaining high accuracy, leading to significant improvements in inference efficiency.
  arXiv  Detail & Related papers  (2024-11-21T14:22:38Z)
• Large Vision-Language Models as Emotion Recognizers in Context Awareness [14.85890824622433]
  Context-aware emotion recognition (CAER) is a complex and significant task that requires perceiving emotions from various contextual cues. Previous approaches primarily focus on designing sophisticated architectures to extract emotional cues from images. This paper systematically explore the potential of leveraging Large Vision-Language Models (LVLMs) to empower the CAER task.
  arXiv  Detail & Related papers  (2024-07-16T01:28:06Z)
• Chain-of-Spot: Interactive Reasoning Improves Large Vision-Language Models [81.71651422951074]
  Chain-of-Spot (CoS) method is a novel approach that enhances feature extraction by focusing on key regions of interest. This technique allows LVLMs to access more detailed visual information without altering the original image resolution. Our empirical findings demonstrate a significant improvement in LVLMs' ability to understand and reason about visual content.
  arXiv  Detail & Related papers  (2024-03-19T17:59:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.