paper-with-me

Papers

Window Token Concatenation for Efficient Visual Large Language Models

2025-04-05 · YiFan Li, Wentao Bao, Botao Ye, Zhen Tan, Tianlong Chen, Huan Liu, Yu Kong

To effectively reduce the visual tokens in Visual Large Language Models (VLLMs), we propose a novel approach called Window Token Concatenation (WiCo). Specifically, we employ a sliding window to concatenate spatially adjacent visual tokens. However, directly concatenating these tokens may group diverse tokens into one, and thus obscure some fine details. To address this challenge, we propose fine-tuning the last few layers of the vision encoder to adaptively adjust the visual tokens, encouraging that those within the same window exhibit similar features. To further enhance the performance on fine-grained visual understanding tasks, we introduce WiCo+, which decomposes the visual tokens in later layers of the LLM. Such a design enjoys the merits of the large perception field of the LLM for fine-grained visual understanding while keeping a small number of visual tokens for efficient inference. We perform extensive experiments on both coarse- and fine-grained visual understanding tasks based on LLaVA-1.5 and Shikra, showing better performance compared with existing token reduction projectors. The code is available: https://github.com/JackYFL/WiCo.

📄 PDF Abstract BibTeX arXiv:2504.04024

Code (1)

jackyfl/wico 공식 구현 pytorch

Tasks

Token Reduction

Similar Papers 제목 키워드 기반

Encoding Sentence Position in Context-Aware Neural Machine Translation with Concatenation

2023-02-13 · Lorenzo Lupo, Marco Dinarelli, Laurent Besacier

Context-aware translation can be achieved by processing a concatenation of consecutive sentences with the standard Transformer architecture. This paper investigates the intuitive idea of providing the model with explicit…

Machine TranslationPositionSentenceTranslation

Visual Context Window Extension: A New Perspective for Long Video Understanding

2024-09-30 · Hongchen Wei, Zhenzhong Chen

Large Multimodal Models (LMMs) have demonstrated impressive performance in short video understanding tasks but face great challenges when applied to long video understanding. In contrast, Large Language Models (LLMs) exh…

Video Understanding

B-VLLM: A Vision Large Language Model with Balanced Spatio-Temporal Tokens

2024-12-13 · Zhuqiang Lu, Zhenfei Yin, Mengwei He, Zhihui Wang 외

Recently, Vision Large Language Models (VLLMs) integrated with vision encoders have shown promising performance in vision understanding. The key of VLLMs is to encode visual content into sequences of visual tokens, enabl…

Language ModelingLanguage ModellingLarge Language ModelVideo Understanding

WindowQuant: Mixed-Precision KV Cache Quantization based on Window-Level Similarity for VLMs Inference Optimization

2026-05-04 · Wei Tao, Xiaoyang Qu, Peiqiang Wang, Guokuan Li 외 arxiv

Recently, video language models (VLMs) have been applied in various fields. However, the visual token sequence of the VLM is too long, which may cause intolerant inference latency and GPU memory usage. Existing methods p…

Sparrow: Text-Anchored Window Attention with Visual-Semantic Glimpsing for Speculative Decoding in Video LLMs

2026-02-17 · Libo Zhang, Zhaoning Zhang, Wangyang Hong, Peng Qiao 외 arxiv

Although speculative decoding is widely used to accelerate Vision-Language Models (VLMs) inference, it faces severe performance collapse when applied to Video Large Language Models (Vid-LLMs). The draft model typically f…