Visual Token Compression

Follow this topic Ask about this topic

Vision encoders represent images and video as patch or frame tokens, and long token sequences can make multimodal models costly to run, especially during prefill and memory-intensive video reasoning. Methods include training-free pruning and merging, learned resamplers, quantization and codecs, and adaptive token budgets guided by text queries, spatial structure, or temporal redundancy. They are evaluated on tasks such as visual question answering, video understanding, OCR, segmentation, retrieval, generation, and robot control. The central tension is that fewer tokens do not always mean faster end-to-end inference, and aggressive compression can discard details needed for a particular question or task.

Latest papers

271 papers46 papers in the last four weeks, up 229% on the four weeks before. 0.3% of all new papers.

Apr 20, 2026

Geometry-Guided 3D Visual Token Pruning for Video-Language Models

Multimodal large language models have demonstrated remarkable capabilities in 2D vision, motivating their extension to 3D scene understanding. Recent studies represent 3D scenes as 3D spatial videos composed of image sequences with depth and camera pose information, enabling pre-trained video-language models to perform 3D reasoning tasks. However, the large number of visual tokens in spatial videos remains a major bottleneck for efficient inference and context management. Existing pruning methods overlook the view consistency of spatial videos and the spatial diversity of the remaining tokens, which prevents them from effectively removing inter-frame redundancy and preserving scene completeness. In this paper, we propose Geo3DPruner, a Geometry-Guided 3D Visual Token Pruning framework. Geo3DPruner first models cross-frame relevance through geometry-aware global attention, and then performs a two-stage pruning process. The intra-voxel stage selects representative multi-view features within each voxel, while the inter-voxel stage preserves spatial diversity by selecting a globally distributed subset of voxels. Extensive experiments on multiple 3D scene understanding benchmarks demonstrate that Geo3DPruner retains over 90% of the original performance while pruning 90% of visual tokens, significantly outperforming existing text-guided and vision-guided pruning methods.

Han Li, Zehao Huang, Jiahui Fu +2

Beihang University · Zhongguancun Academy

Visual Token Pruning Visual Token Compression 3D Scene Understanding 3D VLMs

Apr 18, 2026

EvoComp: Learning Visual Token Compression for Multimodal Large Language Models via Semantic-Guided Evolutionary Labeling

Recent Multimodal Large Language Models (MLLMs) have demonstrated strong performance on vision-language understanding tasks, yet their inference efficiency is often hampered by the large number of visual tokens, particularly in high-resolution or multi-image scenarios. To address this issue, we propose EvoComp, a visual token compression framework that significantly reduces token count while preserving task accuracy. EvoComp introduces a lightweight encoder-only transformer-based compressor that selects the most informative and non-redundant visual tokens by jointly considering visual and textual contexts. A core challenge lies in providing effective supervision for training the compressor. To this end, we design an evolutionary labeling strategy that searches for token subsets minimizing the MLLM's output loss, while enforcing semantic diversity through vocabulary-based token grouping. We further train the compressor using a tailored loss function combining the GHM loss to mitigate class and difficulty imbalance, and a cosine similarity regularization to encourage semantic separation between retained and discarded tokens. Extensive experiments across multiple vision-language benchmarks show that EvoComp outperforms existing methods based on attention or similarity heuristics. Notably, it retains 99.3% of the original accuracy under 3x token compression and delivers up to 1.6x speedup on mobile devices.

Jiafei Song, Fengwei Zhou, Jin Qu +7

OPPO CTG

Efficient Multimodal Inference Multimodal Large Language Models Multimodal Token Compression Visual Token Compression

Apr 17, 2026

Why Training-Free Token Reduction Collapses: The Inherent Instability of Pairwise Scoring Signals

Training-free token reduction methods for Vision Transformers (ToMe, ToFu, PiToMe, and MCTF) employ different scoring mechanisms, yet they share a closely matched cliff-like collapse at high compression. This paper explains why. We develop a diagnostic framework with two tools, ranking consistency

ρ_s

and off-diagonal correlation

ρ_\text{off}

, that decomposes the collapse into (1)a signal-agnostic error amplifier inherent to layer-wise reduction, predicting convex Pareto curves and

r_{\text{crit}} \propto 1/L

; and (2)shared reliance on pairwise similarity signals whose ranking consistency degrades from

ρ_s{=}0.88

to

0.27

in deep layers. Pairwise rankings are inherently unstable (

O(N_p^2)

joint perturbations) while unary signals enjoy greater stability (

O(N_p)

perturbations, CLT). From three design principles derived from this diagnosis, we construct CATIS as a constructive validation: unary signals raise the trigger threshold, triage suppresses the gain. On ViT-Large at 63% FLOPs reduction, CATIS retains 96.9% of vanilla accuracy (81.0%) on ImageNet-1K where all baselines collapse to 43--65%.

Yang Shanglin

Visual Token Pruning Visual Token Compression

Apr 16, 2026

StreamCacheVGGT: Streaming Visual Geometry Transformers with Robust Scoring and Hybrid Cache Compression

Reconstructing dense 3D geometry from continuous video streams requires stable inference under a constant memory budget. Existing

O(1)

frameworks primarily rely on a ``pure eviction'' paradigm, which suffers from significant information destruction due to binary token deletion and evaluation noise from localized, single-layer scoring. To address these bottlenecks, we propose StreamCacheVGGT, a training-free framework that reimagines cache management through two synergistic modules: Cross-Layer Consistency-Enhanced Scoring (CLCES) and Hybrid Cache Compression (HCC). CLCES mitigates activation noise by tracking token importance trajectories across the Transformer hierarchy, employing order-statistical analysis to identify sustained geometric salience. Leveraging these robust scores, HCC transcends simple eviction by introducing a three-tier triage strategy that merges moderately important tokens into retained anchors via nearest-neighbor assignment on the key-vector manifold. This approach preserves essential geometric context that would otherwise be lost. Extensive evaluations on five benchmarks (7-Scenes, NRGBD, ETH3D, Bonn, and KITTI) demonstrate that StreamCacheVGGT sets a new state-of-the-art, delivering superior reconstruction accuracy and long-term stability while strictly adhering to constant-cost constraints.

Xuanyi Liu, Chunan Yu, Deyi Ji +6

KOKONI 3D, Moxin Technology · Nanjing University of Science and Technology · University of Science and Technology of China +4

3D ViTs Streaming 3D Reconstruction KV Caching 3D Reconstruction Efficient ViTs Visual Token Compression+1

Apr 16, 2026

Revisiting Token Compression for Accelerating ViT-based Sparse Multi-View 3D Object Detectors

Vision Transformer (ViT)-based sparse multi-view 3D object detectors have achieved remarkable accuracy but still suffer from high inference latency due to heavy token processing. To accelerate these models, token compression has been widely explored. However, our revisit of existing strategies, such as token pruning, merging, and patch size enlargement, reveals that they often discard informative background cues, disrupt contextual consistency, and lose fine-grained semantics, negatively affecting 3D detection. To overcome these limitations, we propose SEPatch3D, a novel framework that dynamically adjusts patch sizes while preserving critical semantic information within coarse patches. Specifically, we design Spatiotemporal-aware Patch Size Selection (SPSS) that assigns small patches to scenes containing nearby objects to preserve fine details and large patches to background-dominated scenes to reduce computation cost. To further mitigate potential detail loss, Informative Patch Selection (IPS) selects the informative patches for feature refinement, and Cross-Granularity Feature Enhancement (CGFE) injects fine-grained details into selected coarse patches, enriching semantic features. Experiments on the nuScenes and Argoverse 2 validation sets show that SEPatch3D achieves up to 57% faster inference than the StreamPETR baseline and 20% higher efficiency than the state-of-the-art ToC3D-faster, while preserving comparable detection accuracy. Code is available at https://github.com/Mingqj/SEPatch3D.

Mingqian Ji, Shanshan Zhang, Jian Yang

PCA Lab, School of Computer Science and Engineering, Nanjing University of Science and Technology · PCA Lab, School of Intelligence Science and Technology, Nanjing University

Multi-View 3D Object Detection Efficient ViTs 3D Object Detection Visual Token Compression

Apr 15, 2026

UHR-BAT: Budget-Aware Token Compression Vision-Language model for Ultra-High-Resolution Remote Sensing

Ultra-high-resolution (UHR) remote sensing imagery couples kilometer-scale context with query-critical evidence that may occupy only a few pixels. Such vast spatial scale leads to a quadratic explosion of visual tokens and hinders the extraction of information from small objects. Previous works utilize direct downsampling, dense tiling, or global top-k pruning, which either compromise query-critical image details or incur unpredictable compute. In this paper, we propose UHR-BAT, a query-guided and region-faithful token compression framework to efficiently select visual tokens under a strict context budget. Specifically, we leverage text-guided, multi-scale importance estimation for visual tokens, effectively tackling the challenge of achieving precise yet low-cost feature extraction. Furthermore, by introducing region-wise preserve and merge strategies, we mitigate visual token redundancy, further driving down the computational budget. Experimental results show that UHR-BAT achieves state-of-the-art performance across various benchmarks. Code will be available at https://github.com/Yunkaidang/UHR.

Yunkai Dang, Minxin Dai, Yuekun Yang +4

School of Artificial Intelligence Science and Technology, Nanjing University · School of Electronic Science and Engineering, Nanjing University · School of Physics, Nanjing University

Visual Token Compression Remote Sensing Image Understanding Large Vision-Language Models Visual Token Merging

Apr 15, 2026

MaMe & MaRe: Matrix-Based Token Merging and Restoration for Efficient Visual Perception and Synthesis

Token compression is crucial for mitigating the quadratic complexity of self-attention mechanisms in Vision Transformers (ViTs), which often involve numerous input tokens. Existing methods, such as ToMe, rely on GPU-inefficient operations (e.g., sorting, scattered writes), introducing overheads that limit their effectiveness. We introduce MaMe, a training-free, differentiable token merging method based entirely on matrix operations, which is GPU-friendly to accelerate ViTs. Additionally, we present MaRe, its inverse operation, for token restoration, forming a MaMe+MaRe pipeline for image synthesis. When applied to pre-trained models, MaMe doubles ViT-B throughput with a 2% accuracy drop. Notably, fine-tuning the last layer with MaMe boosts ViT-B accuracy by 1.0% at 1.1x speed. In SigLIP2-B@512 zero-shot classification, MaMe provides 1.3x acceleration with negligible performance degradation. In video tasks, MaMe accelerates VideoMAE-L by 48.5% on Kinetics-400 with only a 0.84% accuracy loss. Furthermore, MaMe achieves simultaneous improvements in both performance and speed on some tasks. In image synthesis, the MaMe+MaRe pipeline enhances quality while reducing Stable Diffusion v2.1 generation latency by 31%. Collectively, these results demonstrate MaMe's and MaRe's effectiveness in accelerating vision models. The code is available at https://github.com/cominder/mame}{https://github.com/cominder/mame.

Simin Huo, Ning Li

Shanghai Jiao Tong University, , Shanghai, 200240, Shanghai, China

Visual Token Merging Visual Token Compression Efficient ViTs Diffusion Model Acceleration

Apr 13, 2026

POINTS-Long: Adaptive Dual-Mode Visual Reasoning in MLLMs

Multimodal Large Language Models (MLLMs) have recently demonstrated remarkable capabilities in cross-modal understanding and generation. However, the rapid growth of visual token sequences--especially in long-video and streaming scenarios--poses a major challenge to their scalability and real-world deployment. Thus, we introduce POINTS-Long, a native dual-mode MLLM featuring dynamic visual token scaling inspired by the human visual system. The model supports two complementary perception modes: focus mode and standby mode, enabling users to dynamically trade off efficiency and accuracy during inference. On fine-grained visual tasks, the focus mode retains the optimal performance, while on long-form general visual understanding, the standby mode retains 97.7-99.7% of the original accuracy using only 1/40-1/10th of the visual tokens. Moreover, POINTS-Long natively supports streaming visual understanding via a dynamically detachable KV-cache design, allowing efficient maintenance of ultra-long visual memory. Our work provides new insights into the design of future MLLMs and lays the foundation for adaptive and efficient long-form visual understanding.

Haicheng Wang, Yuan Liu, Yikun Liu +9

SAI, Shanghai Jiao Tong University, China · WeChat AI, Tencent, China

Multimodal Large Language Models Long-Context VLMs Efficient VLM Inference Visual Token Compression Memory-Augmented VLMs

Apr 13, 2026

Semantic-Geometric Dual Compression: Training-Free Visual Token Reduction for Ultra-High-Resolution Remote Sensing Understanding

Multimodal Large Language Models (MLLMs) have demonstrated immense potential in Earth observation. However, the massive visual tokens generated when processing Ultra-High-Resolution (UHR) imagery introduce prohibitive computational overhead, severely bottlenecking their inference efficiency. Existing visual token compression methods predominantly adopt static and uniform compression strategies, neglecting the inherent "Semantic-Geometric Duality" in remote sensing interpretation tasks. Specifically, object semantic tasks focus on the abstract semantics of objects and benefit from aggressive background pruning, whereas scene geometric tasks critically rely on the integrity of spatial topology. To address this challenge, we propose DualComp, a task-adaptive dual-stream token compression framework. Dynamically guided by a lightweight pre-trained router, DualComp decouples feature processing into two dedicated pathways. In the object semantic stream, the Spatially-Contiguous Semantic Aggregator (SCSA) utilizes size-adaptive clustering to aggregates redundant background while protecting small object. In the scene geometric stream, the Instruction-Guided Structure Recoverer (IGSR) introduces a greedy path-tracing topology completion mechanism to reconstruct spatial skeletons. Experiments on the UHR remote sensing benchmark XLRS-Bench demonstrate that DualComp accomplishes high-fidelity remote sensing interpretation at an exceptionally low computational cost, achieving simultaneous improvements in both efficiency and accuracy.

Yueying Li, Fengxiang Wang, Yan Li +3

College of Computer Science and Technology, National University of Defense Technology, Changsha 410073, China · School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876, China

Visual Token Compression Efficient VLM Inference Remote Sensing Image Understanding

Apr 11, 2026

Visual Late Chunking: An Empirical Study of Contextual Chunking for Efficient Visual Document Retrieval

Multi-vector models dominate Visual Document Retrieval (VDR) due to their fine-grained matching capabilities, but their high storage and computational costs present a major barrier to practical deployment. In this paper, we propose ColChunk, a plug-and-play framework that introduces multimodal late chunking to construct efficient, contextualized multi-vectors. Unlike existing pruning or fixed-token approaches, ColChunk employs hierarchical clustering on patch-level embeddings, fused with a 2D position prior to ensure spatial-semantic coherence. This adaptive grouping allows for a content-aware representation that preserves global context while drastically reducing the vector count. Evaluations across 24 VDR datasets demonstrate ColChunk achieves over a 90% reduction in storage requirements while simultaneously delivering a 9-point average improvement in nDCG@5 across representative single-vector models. ColChunk provides a practical solution for balancing retrieval accuracy and efficiency in visual document systems.

Yibo Yan, Mingdong Ou, Yi Cao +5

Hong Kong University of Science and Technology (Guangzhou) · Alibaba Cloud Computing · Hong Kong University of Science and Technology

Visual Document Retrieval Multi-Vector Retrieval Document Chunking Visual Token Compression

Apr 10, 2026

UIPress: Bringing Optical Token Compression to UI-to-Code Generation

UI-to-Code generation requires vision-language models (VLMs) to produce thousands of tokens of structured HTML/CSS from a single screenshot, making visual token efficiency critical. Existing compression methods either select tokens at inference time using task-agnostic heuristics, or zero out low-attention features without actually shortening the sequence -- neither truly reduces prefill latency or adapts to the non-uniform information density of UI screenshots. Meanwhile, optical (encoder-side learned) compression has shown strong results for document OCR, yet no prior work has adapted this paradigm to UI-to-Code generation. We propose UIPress, a lightweight learned compression module inserted between the frozen ViT encoder and the LLM decoder of Qwen3-VL-8B. UIPress combines depthwise-separable convolutions, element-guided spatial reweighting, and Transformer refinement to compress

{\sim}

6{,}700 visual tokens to a fixed budget of 256. Together with Low-Rank Adaptation (LoRA) on the decoder to bridge the representation gap, the entire system adds only

{\sim}

21.7M trainable parameters (0.26% of the 8B base model). Under a fair comparison on the same base model against four baselines on Design2Code, UIPress at 256 tokens achieves a CLIP score of 0.8127, outperforming the uncompressed baseline by +7.5% and the strongest inference-time method by +4.6%, while delivering 9.1

\times

time-to-first-token speedup. To the best of our knowledge, UIPress is the first encoder-side learned compression method for the UI-to-Code task.

Dasen Dai, Shuoqi Li, Ronghao Chen +3

The Chinese University of Hong Kong China · Peking University China · University of Chinese Academy of Sciences China +2

Token Compression Visual Token Compression Image-to-Code Generation Efficient VLM Inference Code Generation Vision-Language Models

Apr 10, 2026

Do Vision Language Models Need to Process Image Tokens?

Vision Language Models (VLMs) have achieved remarkable success by integrating visual encoders with large language models (LLMs). While VLMs process dense image tokens across deep transformer stacks (incurring substantial computational overhead), it remains fundamentally unclear whether sustained image-token processing is necessary for their performance or visual representations meaningfully evolve from early to later layers. In this work, we systematically investigate the functional role of image tokens in VLMs and show that visual representations rapidly converge to a bounded-complexity regime, \ie their entropy stabilizes, intrinsic dimensionality compresses, and trajectory curvature approaches a near-constant profile. In contrast, textual representations continue to undergo substantial restructuring across depth. Once stabilized, visual representations become largely interchangeable between layers, indicating limited additional transformation in deeper stages. Further, depth-wise visual truncation reveals that the necessity of visual processing is task-dependent, where single-token predictions remain comparatively robust to truncated visual depth, but multi-token generation require sustained access to visual representations. Under deterministic decoding, reducing visual depth perturbs intermediate reasoning trajectories more strongly than final outputs, suggesting that image tokens influence the structure of reasoning more than the ultimate conclusions. Collectively, these findings question the assumption that deeper visual processing is uniformly essential in VLMs, challenging the current paradigm of multimodal LLM architectures.

Sambit Ghosh, R. Venkatesh Babu, Chirag Agarwal

IBM · Indian Institute of Science · University of Virginia

Vision-Language Models VLM Reasoning Visual Token Compression

Apr 9, 2026

EPIR: An Efficient Patch Tokenization, Integration and Representation Framework for Micro-expression Recognition

Micro-expression recognition can obtain the real emotion of the individual at the current moment. Although deep learning-based methods, especially Transformer-based methods, have achieved impressive results, these methods have high computational complexity due to the large number of tokens in the multi-head self-attention. In addition, the existing micro-expression datasets are small-scale, which makes it difficult for Transformer-based models to learn effective micro-expression representations. Therefore, we propose a novel Efficient Patch tokenization, Integration and Representation framework (EPIR), which can balance high recognition performance and low computational complexity. Specifically, we first propose a dual norm shifted tokenization (DNSPT) module to learn the spatial relationship between neighboring pixels in the face region, which is implemented by a refined spatial transformation and dual norm projection. Then, we propose a token integration module to integrate partial tokens among multiple cascaded Transformer blocks, thereby reducing the number of tokens without information loss. Furthermore, we design a discriminative token extractor, which first improves the attention in the Transformer block to reduce the unnecessary focus of the attention calculation on self-tokens, and uses the dynamic token selection module (DTSM) to select key tokens, thereby capturing more discriminative micro-expression representations. We conduct extensive experiments on four popular public datasets (i.e., CASME II, SAMM, SMIC, and CAS(ME)3. The experimental results show that our method achieves significant performance gains over the state-of-the-art methods, such as 9.6% improvement on the CAS(ME)

^3

dataset in terms of UF1 and 4.58% improvement on the SMIC dataset in terms of UAR metric.

Junbo Wang, Liangyu Fu, Yuke Li +3

School of Software, Northwestern Polytechnical University, Xi’an, 710129, China · School of Computer Science, Northwestern Polytechnical University, Xi’an, 710129, China · School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, 710049, China +1

Micro-Expression Recognition Facial Expression Recognition Efficient ViTs Visual Token Compression Visual Token Pruning

Apr 8, 2026

TC-AE: Unlocking Token Capacity for Deep Compression Autoencoders

We propose TC-AE, a ViT-based architecture for deep compression autoencoders. Existing methods commonly increase the channel number of latent representations to maintain reconstruction quality under high compression ratios. However, this strategy often leads to latent representation collapse, which degrades generative performance. Instead of relying on increasingly complex architectures or multi-stage training schemes, TC-AE addresses this challenge from the perspective of the token space, the key bridge between pixels and image latents, through two complementary innovations: Firstly, we study token number scaling by adjusting the patch size in ViT under a fixed latent budget, and identify aggressive token-to-latent compression as the key factor that limits effective scaling. To address this issue, we decompose token-to-latent compression into two stages, reducing structural information loss and enabling effective token number scaling for generation. Secondly, to further mitigate latent representation collapse, we enhance the semantic structure of image tokens via joint self-supervised training, leading to more generative-friendly latents. With these designs, TC-AE achieves substantially improved reconstruction and generative performance under deep compression. We hope our research will advance ViT-based tokenizer for visual generation.

Teng Li, Ziyuan Huang, Cong Chen +5

Inclusion AI, Ant Group · HKUST · ZJU +2

Autoencoders Visual Token Compression Visual Tokenization Vision Transformer

Apr 6, 2026

Hierarchical SVG Tokenization: Learning Compact Visual Programs for Scalable Vector Graphics Modeling

Recent large language models have shifted SVG generation from differentiable rendering optimization to autoregressive program synthesis. However, existing approaches still rely on generic byte-level tokenization inherited from natural language processing, which poorly reflects the geometric structure of vector graphics. Numerical coordinates are fragmented into discrete symbols, destroying spatial relationships and introducing severe token redundancy, often leading to coordinate hallucination and inefficient long-sequence generation. To address these challenges, we propose HiVG, a hierarchical SVG tokenization framework tailored for autoregressive vector graphics generation. HiVG decomposes raw SVG strings into structured atomic tokens and further compresses executable command--parameter groups into geometry-constrained segment tokens, substantially improving sequence efficiency while preserving syntactic validity. To further mitigate spatial mismatch, we introduce a Hierarchical Mean--Noise (HMN) initialization strategy that injects numerical ordering signals and semantic priors into new token embeddings. Combined with a curriculum training paradigm that progressively increases program complexity, HiVG enables more stable learning of executable SVG programs. Extensive experiments on both text-to-SVG and image-to-SVG tasks demonstrate improved generation fidelity, spatial consistency, and sequence efficiency compared with conventional tokenization schemes. Our code is publicly available at https://github.com/ximinng/HiVG

Ximing Xing, Ziteng Xue, Zhenxi Li +8

Tencent · Visual Computing Group, Beijing

Vector Graphics Generation Visual Tokenization Text-to-SVG Generation Visual Token Compression

Apr 4, 2026

RCP: Representation Consistency Pruner for Mitigating Distribution Shift in Large Vision-Language Models

Large Vision-Language Models (LVLMs) suffer from prohibitive inference costs due to the massive number of visual tokens processed by the language decoder. Existing pruning methods often lead to significant performance degradation because the irreversible removal of visual tokens causes a distribution shift in the hidden states that deviates from the pre-trained full-token regime. To address this, we propose Representation Consistency Pruner, which we refer to as RCP, as a novel framework that integrates cumulative visual token pruning with a delayed repair mechanism. Specifically, we introduce a cross-attention pruner that leverages the intrinsic attention of the LLM as a baseline to predict cumulative masks, ensuring consistent and monotonic token reduction across layers. To compensate for the resulting information loss, we design a delayed repair adapter denoted as DRA, which caches the essence of pruned tokens and applies FiLM-based modulation specifically to the answer generation tokens. We employ a repair loss to match the first and second-order statistics of the pruned representations with a full-token teacher. RCP is highly efficient because it trains only lightweight plug-in modules while allowing for physical token discarding at inference. Extensive experiments on LVLM benchmarks demonstrate that RCP removes up to 88.9% of visual tokens and reduces FLOPs by up to 85.7% with only a marginal average accuracy drop, and outperforms prior methods that avoid fine-tuning the original model on several widely used benchmarks.

Jianwei Zhang, Chaoning Zhang, Sihan Cao +7

School of Computer Science and Engineering, University of Electronic Science and Technology of China, Qingshuihe Campus, No. 2006 Xiyuan Avenue, High-Tech Zone (West), Chengdu 611731, China · Machine Learning Department, Mohamed bin Zayed University of Artificial Intelligence, Building 1B, Masdar City, Abu Dhabi, United Arab Emirates · School of Robotics and Advanced Manufacture, Harbin Institute of Technology, Shenzhen, University Town of Shenzhen, Nanshan District, Shenzhen 518055, China +1

Efficient VLM Inference Large Vision-Language Models Visual Token Pruning Visual Token Compression

Apr 3, 2026

KiToke: Distribution-Aware Token Compression for Video Large Language Models

Video Large Language Models (Video LLMs) achieve strong performance on video understanding tasks but suffer from high inference costs due to the large number of visual tokens. We study training-free, query-agnostic video token compression, where heuristic importance rules can over-represent large redundant visual groups and miss less frequent visual patterns, particularly under aggressive token budgets. We propose KiToke, a distribution-aware compression framework that treats visual tokens as samples from the video's spatiotemporal token distribution. KiToke uses global embedding density as a proxy for representational redundancy and inverse-density sampling to balance token allocation across visual patterns of different frequencies. It further aggregates unselected tokens into retained anchors within detected temporal intervals to reduce cross-transition mixing. Across evaluated backbones, KiToke achieves the highest average accuracy among compared training-free methods and remains robust down to 1% retention, where most evaluated baselines underperform uniform random token sampling. On LLaVA-OneVision, it preserves 98.5% and 89.2% of the uncompressed model's four-benchmark mean accuracy at 10% and 1% token retention, with 6.8x and 12.5x prefill speedups, respectively, including compression overhead.

Haifeng Huang, Biyin Xu, Yang Li

Department of Computer Science, Iowa State University.

Video Token Compression Visual Token Compression Efficient VLM Inference Video-Language Models Visual Token Merging

Apr 1, 2026

PixelPrune: Pixel-Level Adaptive Visual Token Reduction via Predictive Coding

Document understanding and GUI interaction are among the highest-value applications of Vision-Language Models (VLMs), yet they impose exceptionally heavy computational burden: fine-grained text and small UI elements demand high-resolution inputs that produce tens of thousands of visual tokens. We observe that this cost is largely wasteful -- across document and GUI benchmarks, only 22--71% of image patches are pixel-unique, the rest being exact duplicates of another patch in the same image. We propose PixelPrune, which exploits this pixel-level redundancy through predictive-coding-based compression, pruning redundant patches before the Vision Transformer (ViT) encoder. Because it operates in pixel space prior to any neural computation, PixelPrune accelerates both the ViT encoder and the downstream LLM, covering the full inference pipeline. The method is training-free, requires no learnable parameters, and supports pixel-lossless compression (

τ{=}0

) as well as controlled lossy compression (

τ{>}0

). Experiments across three model scales and document and GUI benchmarks show that PixelPrune maintains competitive task accuracy while delivering up to 4.2

\times

inference speedup and 1.9

\times

training acceleration. Code is available at https://github.com/OPPO-Mente-Lab/PixelPrune.

Nan Wang, Zhiwei Jin, Chen Chen +1

OPPO AI Center

Efficient VLM Inference Visual Token Pruning Visual Token Compression

Mar 31, 2026

MacTok: Robust Continuous Tokenization for Image Generation

Continuous image tokenizers enable efficient visual generation, and those based on variational frameworks can learn smooth, structured latent representations through KL regularization. Yet this often leads to posterior collapse when using fewer tokens, where the encoder fails to encode informative features into the compressed latent space. To address this, we introduce MacTok, a Masked Augmenting 1D Continuous Tokenizer that leverages image masking and representation alignment to prevent collapse while learning compact and robust representations. MacTok applies both random masking to regularize latent learning and DINO-guided semantic masking to emphasize informative regions in images, forcing the model to encode robust semantics from incomplete visual evidence. Combined with global and local representation alignment, MacTok preserves rich discriminative information in a highly compressed 1D latent space, requiring only 64 or 128 tokens. On ImageNet, MacTok achieves a competitive gFID of 1.44 at 256

\times

256 and a state-of-the-art 1.52 at 512

\times

512 with SiT-XL, while reducing token usage by up to 64

\times

. These results confirm that masking and semantic guidance together prevent posterior collapse and achieve efficient, high-fidelity tokenization.

Hengyu Zeng, Xin Gao, Guanghao Li +5

Fudan University

Image Tokenization Visual Token Compression Posterior Collapse

Mar 30, 2026

Towards Efficient Large Vision-Language Models: A Comprehensive Survey on Inference Strategies

Although Large Vision Language Models (LVLMs) have demonstrated impressive multimodal reasoning capabilities, their scalability and deployment are constrained by massive computational requirements. In particular, the massive amount of visual tokens from high-resolution input data aggravates the situation due to the quadratic complexity of attention mechanisms. To address these issues, the research community has developed several optimization frameworks. This paper presents a comprehensive survey of the current state-of-the-art techniques for accelerating LVLM inference. We introduce a systematic taxonomy that categorizes existing optimization frameworks into four primary dimensions: visual token compression, memory management and serving, efficient architectural design, and advanced decoding strategies. Furthermore, we critically examine the limitations of these current methodologies and identify critical open problems to inspire future research directions in efficient multimodal systems.

Surendra Pathak, Bo Han

Computer Science Department, George Mason University, Fairfax, VA - 22030, USA

Efficient VLM Inference Visual Token Compression

Mar 26, 2026

Semantic-Aware Prefix Learning for Token-Efficient Image Generation

Visual tokenizers play a central role in latent image generation by bridging high-dimensional images and tractable generative modeling. However, most existing tokenizers are still trained with reconstruction-dominated objectives, which often yield latent representations that are only weakly grounded in high-level semantics. Recent approaches improve semantic alignment, but typically treat semantic signals as auxiliary regularization rather than making them functionally necessary for representation learning. We propose SMAP, a SeMantic-Aware Prefix tokenizer that injects class-level semantic conditions into a query-based 1D tokenization framework. To make semantics indispensable during training, SMAP introduces a tail token dropping strategy, which forces semantic conditions and early latent prefixes to bear increasing responsibility under progressively reduced token budgets. To verify that the resulting latent space is useful for generation rather than reconstruction alone, we further introduce CARD, a hybrid Causal AutoRegressive--Diffusion generator. Extensive experiments on ImageNet show that SMAP consistently improves reconstruction quality across discrete and continuous tokenization settings, and that its semantically grounded latent space yields strong downstream generation performance under compact token budgets.

Qingfeng Li, Haoxian Zhang, Xu He +4

Institute of Automation, Chinese Academy of Sciences · Kling Team, Kuaishou Technology · Tsinghua University

Visual Tokenization Visual Token Compression Conditional Image Generation Autoregressive Image Generation

Mar 26, 2026

Photon: Speedup Volume Understanding with Efficient Multimodal Large Language Models

Multimodal large language models are promising for clinical visual question answering tasks, but scaling to 3D imaging is hindered by high computational costs. Prior methods often rely on 2D slices or fixed-length token compression, disrupting volumetric continuity and obscuring subtle findings. We present Photon, a framework that represents 3D medical volumes with token sequences of variable length. Photon introduces instruction-conditioned token scheduling and surrogate gradient propagation to adaptively reduce tokens during both training and inference, which lowers computational cost while mitigating the attention dilution caused by redundant tokens. It incorporates a custom backpropagation rule with gradient restoration to enable differentiable optimization despite discrete token drop. To stabilize token compression and ensure reliable use of visual evidence, Photon further applies regularization objectives that mitigate language-only bias and improve reliability. Experiments on diverse medical visual question answering tasks show that Photon achieves state-of-the-art accuracy while reducing resource usage and accelerating both training and inference.

Chengyu Fang, Heng Guo, Zheng Jiang +3

Tsinghua University · DAMO Academy, Alibaba Group · Hupan Lab +1

Medical VLMs 3D VLMs Efficient Multimodal Inference Medical VQA Visual Token Compression

Mar 23, 2026

Rethinking Token Reduction for Large Vision-Language Models

Large Vision-Language Models (LVLMs) excel in visual understanding and reasoning, but the excessive visual tokens lead to high inference costs. Although recent token reduction methods mitigate this issue, they mainly target single-turn Visual Question Answering (VQA), leaving the more practical multi-turn VQA (MT-VQA) scenario largely unexplored. MT-VQA introduces additional challenges, as subsequent questions are unknown beforehand and may refer to arbitrary image regions, making existing reduction strategies ineffective. Specifically, current approaches fall into two categories: prompt-dependent methods, which bias toward the initial text prompt and discard information useful for subsequent turns; prompt-agnostic ones, which, though technically applicable to multi-turn settings, rely on heuristic reduction metrics such as attention scores, leading to suboptimal performance. In this paper, we propose a learning-based prompt-agnostic method, termed MetaCompress, overcoming the limitations of heuristic designs. We begin by formulating token reduction as a learnable compression mapping, unifying existing formats such as pruning and merging into a single learning objective. Upon this formulation, we introduce a data-efficient training paradigm capable of learning optimal compression mappings with limited computational costs. Extensive experiments on MT-VQA benchmarks and across multiple LVLM architectures demonstrate that MetaCompress achieves superior efficiency-accuracy trade-offs while maintaining strong generalization across dialogue turns. Our code is available at https://github.com/MArSha1147/MetaCompress.

Yi Wang, Haofei Zhang, Qihan Huang +7

College of Computer Science and Technology, Zhejiang University · State Key Laboratory of Blockchain and Data Security, Zhejiang University · Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security +3

Efficient VLM Inference Visual Token Compression Large Vision-Language Models Visual Token Pruning Visual Question Answering

Mar 22, 2026

QMoP: Query Guided Mixture-of-Projector for Efficient Visual Token Compression

Multimodal large language models suffer from severe computational and memory bottlenecks, as the number of visual tokens far exceeds that of textual tokens. While recent methods employ projector modules to align and compress visual tokens into text-aligned features, they typically depend on fixed heuristics that limit adaptability across diverse scenarios. In this paper, we first propose Query Guided Mixture-of-Projector (QMoP), a novel and flexible framework that adaptively compresses visual tokens via three collaborative branches: (1) a pooling-based branch for coarse-grained global semantics, (2) a resampler branch for extracting high-level semantic representations, and (3) a pruning-based branch for fine-grained token selection to preserve critical visual detail. To adaptively coordinate these branches, we introduce the Query Guided Router (QGR), which dynamically selects and weights the outputs from different branches based on both visual input and textual queries. A Mixture-of-Experts-style fusion mechanism is designed to aggregate the outputs, harnessing the strengths of each strategy while suppressing noise. To systematically evaluate the effects of Visual Token Compression, we also develop VTCBench, a dedicated benchmark for evaluating the information loss induced by visual token compression. Extensive experiments demonstrate that despite relying on fundamental compression modules, QMoP outperforms strong baselines and delivers significant savings in memory, computation, and inference time.

Zhongyang Li, Yaqian Li, Faming Fang +6

East China Normal University · Li Auto Inc.

Multimodal Token Compression Visual Token Compression Efficient VLM Inference Visual Token Pruning

Mar 20, 2026

CurveStream: Boosting Streaming Video Understanding in MLLMs via Curvature-Aware Hierarchical Visual Memory Management

Multimodal Large Language Models have achieved significant success in offline video understanding, yet their application to streaming videos is severely limited by the linear explosion of visual tokens, which often leads to Out-of-Memory (OOM) errors or catastrophic forgetting. Existing visual retention and memory management methods typically rely on uniform sampling, low-level physical metrics, or passive cache eviction. However, these strategies often lack intrinsic semantic awareness, potentially disrupting contextual coherence and blurring transient yet critical semantic transitions. To address these limitations, we propose CurveStream, a training-free, curvature-aware hierarchical visual memory management framework. Our approach is motivated by the key observation that high-curvature regions along continuous feature trajectories closely align with critical global semantic transitions. Based on this geometric insight, CurveStream evaluates real-time semantic intensity via a Curvature Score and integrates an online K-Sigma dynamic threshold to adaptively route frames into clear and fuzzy memory states under a strict token budget. Evaluations across diverse temporal scales confirm that this lightweight framework, CurveStream, consistently yields absolute performance gains of over 10% (e.g., 10.69% on StreamingBench and 13.58% on OVOBench) over respective baselines, establishing new state-of-the-art results for streaming video perception.The code will be released at https://github.com/streamingvideos/CurveStream.

Chao Wang, Xudong Tan, Jianjian Cao +2

College of Future Information Technology, Fudan University, Shanghai, China · Shanghai Innovation Institute, Shanghai, China

Streaming Video Understanding Memory-Augmented Video Understanding Large Vision-Language Models Visual Token Compression

Mar 19, 2026

Multimodal Model for Computational Pathology:Representation Learning and Image Compression

Whole slide imaging (WSI) has transformed digital pathology by enabling computational analysis of gigapixel histopathology images. Recent foundation model advances have accelerated progress in computational pathology, facilitating joint reasoning across pathology images, clinical reports, and structured data. Despite this progress, challenges remain: the extreme resolution of WSIs creates computational hurdles for visual learning; limited expert annotations constrain supervised approaches; integrating multimodal information while preserving biological interpretability remains difficult; and the opacity of modeling ultra-long visual sequences hinders clinical transparency. This review comprehensively surveys recent advances in multimodal computational pathology. We systematically analyze four research directions: (1) self-supervised representation learning and structure-aware token compression for WSIs; (2) multimodal data generation and augmentation; (3) parameter-efficient adaptation and reasoning-enhanced few-shot learning; and (4) multi-agent collaborative reasoning for trustworthy diagnosis. We specifically examine how token compression enables cross-scale modeling and how multi-agent mechanisms simulate a pathologist's "Chain of Thought" across magnifications to achieve uncertainty-aware evidence fusion. Finally, we discuss open challenges and argue that future progress depends on unified multimodal frameworks integrating high-resolution visual data with clinical and biomedical knowledge to support interpretable and safe AI-assisted diagnosis.

Peihang Wu, Zehong Chen, Lijian Xu

Shenzhen University of Advanced Technology

Computational Pathology Whole-Slide Image Analysis Medical Image Representation Learning Multimodal Learning Visual Token Compression Multimodal Model Interpretability

Mar 18, 2026

STAC: Plug-and-Play Spatio-Temporal Aware Cache Compression for Streaming 3D Reconstruction

Online 3D reconstruction from streaming inputs requires both long-term temporal consistency and efficient memory usage. Although causal variants of VGGT address this challenge through a key-value (KV) cache mechanism, the cache grows linearly with the stream length, creating a major memory bottleneck. Under limited memory budgets, early cache eviction significantly degrades reconstruction quality and temporal consistency. In this work, we observe that attention in causal transformers for 3D reconstruction exhibits intrinsic spatio-temporal sparsity. Based on this insight, we propose STAC, a Spatio-Temporally Aware Cache Compression framework for streaming 3D reconstruction with large causal transformers. STAC consists of three key components: (1) a Working Temporal Token Caching mechanism that preserves long-term informative tokens using decayed cumulative attention scores; (2) a Long-term Spatial Token Caching scheme that compresses spatially redundant tokens into voxel-aligned representations for memory-efficient storage; and (3) a Chunk-based Multi-frame Optimization strategy that jointly processes consecutive frames to improve temporal coherence and GPU efficiency. Extensive experiments show that STAC achieves state-of-the-art reconstruction quality while reducing memory consumption by nearly 10x and accelerating inference by 4x, substantially improving the scalability of real-time 3D reconstruction in streaming settings.

Runze Wang, Yuxuan Song, Youcheng Cai +1

University of Science and Technology of China

Streaming 3D Reconstruction KV-Cache Compression 3D ViTs Visual Token Compression

Mar 17, 2026

InViC: Intent-aware Visual Cues for Medical Visual Question Answering

Medical visual question answering (Med-VQA) aims to answer clinically relevant questions grounded in medical images. However, existing multimodal large language models (MLLMs) often exhibit shortcut answering, producing plausible responses by exploiting language priors or dataset biases while insufficiently attending to visual evidence. This behavior undermines clinical reliability, especially when subtle imaging findings are decisive. We propose a lightweight plug-in framework, termed Intent-aware Visual Cues (InViC), to explicitly enhance image-based answer generation in medical VQA. InViC introduces a Cue Tokens Extraction (CTE) module that distills dense visual tokens into a compact set of K question-conditioned cue tokens, which serve as structured visual intermediaries injected into the LLM decoder to promote intent-aligned visual evidence. To discourage bypassing of visual information, we further design a two-stage fine-tuning strategy with a cue-bottleneck attention mask. In Stage I, we employ an attention mask to block the LLM's direct view of raw visual features, thereby funneling all visual evidence through the cue pathway. In Stage II, standard causal attention is restored to train the LLM to jointly exploit the visual and cue tokens. We evaluate InViC on three public Med-VQA benchmarks (VQA-RAD, SLAKE, and ImageCLEF VQA-Med 2019) across multiple representative MLLMs. InViC consistently improves over zero-shot inference and standard LoRA fine-tuning, demonstrating that intent-aware visual cues with bottlenecked training is a practical and effective strategy for improving trustworthy Med-VQA.

Zhisong Wang, Ziyang Chen, Zanting Ye +3

National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, School of Computer Science and Engineering, Northwestern Polytechnical University, Xi’an 710072, China · Southern Medical University · Westlake University

Medical VQA Medical VLMs Visual Token Compression

Mar 16, 2026

Question-guided Visual Compression with Memory Feedback for Long-Term Video Understanding

In the context of long-term video understanding with large multimodal models, many frameworks have been proposed. Although transformer-based visual compressors and memory-augmented approaches are often used to process long videos, they usually compress each frame independently and therefore fail to achieve strong performance on tasks that require understanding complete events, such as temporal ordering tasks in MLVU and VNBench. This motivates us to rethink the conventional one-way scheme from perception to memory, and instead establish a feedbackdriven process in which past visual contexts stored in the context memory can benefit ongoing perception. To this end, we propose Question-guided Visual Compression with Memory Feedback (QViC-MF), a framework for long-term video understanding. At its core is a Question-guided Multimodal Selective Attention (QMSA), which learns to preserve visual information related to the given question from both the current clip and the past related frames from the memory. The compressor and memory feedback work iteratively for each clip of the entire video. This simple yet effective design yields large performance gains on longterm video understanding tasks. Extensive experiments show that our method achieves significant improvement over current state-of-the-art methods by 6.1% on MLVU test, 8.3% on LVBench, 18.3% on VNBench Long, and 3.7% on VideoMME Long. The code will be released publicly.

Sosuke Yamao, Natsuki Miyahara, Yuankai Qi +1

Fujitsu Research · Macquarie University

Long-Video Understanding Temporal Video Understanding Memory-Augmented Video Understanding Video Token Compression Visual Token Compression

Mar 16, 2026

Balancing Saliency and Coverage: Semantic Prominence-Aware Budgeting for Visual Token Compression in VLMs

Large Vision-Language Models (VLMs) achieve strong multimodal understanding capabilities by leveraging high-resolution visual inputs, but the resulting large number of visual tokens creates a major computational bottleneck. Recent work mitigates this issue through visual token compression, typically compressing tokens based on saliency, diversity, or a fixed combination of both. We observe that the distribution of semantic prominence varies substantially across samples, leading to different optimal trade-offs between local saliency preservation and global coverage. This observation suggests that applying a static compression strategy across all samples can be suboptimal. Motivated by this insight, we propose PromPrune, a sample-adaptive visual token selection framework composed of semantic prominence-aware budget allocation and a two-stage selection pipeline. Our method adaptively balances local saliency preservation and global coverage according to the semantic prominence distribution of each sample. By allocating token budgets between locally salient regions and globally diverse regions, our method maintains strong performance even under high compression ratios. On LLaVA-NeXT-7B, our approach reduces FLOPs by 88% and prefill latency by 22% while preserving 97.5% of the original accuracy.

Jaehoon Lee, Mingi Jung, Soohyuk Jang +3

Interdisciplinary Program in AI, Seoul National University · Department of Electrical and Computer Engineering, Seoul National University · Department of Artificial Intelligence, Chung-Ang University

Efficient VLM Inference Visual Token Pruning Visual Token Compression