Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Jinming, Lin, Junyan, Wei, Yuntao, Shao, Kele, Tao, Keda, Huang, Jianguo, Yang, Xudong, Chen, Zhibo, Wang, Huan, Jin, Xin
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2508.13460
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916907722997760
author Liu, Jinming
Lin, Junyan
Wei, Yuntao
Shao, Kele
Tao, Keda
Huang, Jianguo
Yang, Xudong
Chen, Zhibo
Wang, Huan
Jin, Xin
author_facet Liu, Jinming
Lin, Junyan
Wei, Yuntao
Shao, Kele
Tao, Keda
Huang, Jianguo
Yang, Xudong
Chen, Zhibo
Wang, Huan
Jin, Xin
contents Classical visual coding and Multimodal Large Language Model (MLLM) token technology share the core objective - maximizing information fidelity while minimizing computational cost. Therefore, this paper reexamines MLLM token technology, including tokenization, token compression, and token reasoning, through the established principles of long-developed visual coding area. From this perspective, we (1) establish a unified formulation bridging token technology and visual coding, enabling a systematic, module-by-module comparative analysis; (2) synthesize bidirectional insights, exploring how visual coding principles can enhance MLLM token techniques' efficiency and robustness, and conversely, how token technology paradigms can inform the design of next-generation semantic visual codecs; (3) prospect for promising future research directions and critical unsolved challenges. In summary, this study presents the first comprehensive and structured technology comparison of MLLM token and visual coding, paving the way for more efficient multimodal models and more powerful visual codecs simultaneously.
format Preprint
id arxiv_https___arxiv_org_abs_2508_13460
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Revisiting MLLM Token Technology through the Lens of Classical Visual Coding
Liu, Jinming
Lin, Junyan
Wei, Yuntao
Shao, Kele
Tao, Keda
Huang, Jianguo
Yang, Xudong
Chen, Zhibo
Wang, Huan
Jin, Xin
Computer Vision and Pattern Recognition
Classical visual coding and Multimodal Large Language Model (MLLM) token technology share the core objective - maximizing information fidelity while minimizing computational cost. Therefore, this paper reexamines MLLM token technology, including tokenization, token compression, and token reasoning, through the established principles of long-developed visual coding area. From this perspective, we (1) establish a unified formulation bridging token technology and visual coding, enabling a systematic, module-by-module comparative analysis; (2) synthesize bidirectional insights, exploring how visual coding principles can enhance MLLM token techniques' efficiency and robustness, and conversely, how token technology paradigms can inform the design of next-generation semantic visual codecs; (3) prospect for promising future research directions and critical unsolved challenges. In summary, this study presents the first comprehensive and structured technology comparison of MLLM token and visual coding, paving the way for more efficient multimodal models and more powerful visual codecs simultaneously.
title Revisiting MLLM Token Technology through the Lens of Classical Visual Coding
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.13460