Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models

Semantic communication is a promising technology for next-generation wireless networks. However, the out-of-distribution (OOD) problem, where a pre-trained machine learning (ML) model is applied to unseen tasks that are outside the distribution of its training data, may compromise the integrity of s...

Full description

Saved in:
Bibliographic Details
Main Authors: Feifan Zhang, Yuyang Du, Kexin Chen, Yulin Shao, Soung Chang Liew
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Transactions on Machine Learning in Communications and Networking
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11113346/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849738008554635264
author Feifan Zhang
Yuyang Du
Kexin Chen
Yulin Shao
Soung Chang Liew
author_facet Feifan Zhang
Yuyang Du
Kexin Chen
Yulin Shao
Soung Chang Liew
author_sort Feifan Zhang
collection DOAJ
description Semantic communication is a promising technology for next-generation wireless networks. However, the out-of-distribution (OOD) problem, where a pre-trained machine learning (ML) model is applied to unseen tasks that are outside the distribution of its training data, may compromise the integrity of semantic compression. This paper explores the use of multi-modal large language models (MLLMs) to address the OOD issue in image semantic communication. We propose a novel “Plan A - Plan B” framework that leverages the broad knowledge and strong generalization ability of an MLLM to assist a conventional ML model when the latter encounters an OOD input in the semantic encoding process. Furthermore, we propose a Bayesian optimization scheme that reshapes the probability distribution of the MLLM’s inference process based on the contextual information of the image. The optimization scheme significantly enhances the MLLM’s performance in semantic compression by 1) filtering out irrelevant vocabulary in the original MLLM output; and 2) using contextual similarities between prospective answers of the MLLM and the background information as prior knowledge to modify the MLLM’s probability distribution during inference. Further, at the receiver side of the communication system, we put forth a “generate-criticize” framework that utilizes the cooperation of multiple MLLMs to enhance the reliability of image reconstruction.
format Article
id doaj-art-3f23e661038848ab8c36c331b42864dc
institution DOAJ
issn 2831-316X
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Transactions on Machine Learning in Communications and Networking
spelling doaj-art-3f23e661038848ab8c36c331b42864dc2025-08-20T03:06:44ZengIEEEIEEE Transactions on Machine Learning in Communications and Networking2831-316X2025-01-013997101310.1109/TMLCN.2025.359584111113346Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language ModelsFeifan Zhang0https://orcid.org/0009-0009-4153-925XYuyang Du1https://orcid.org/0000-0001-9213-9875Kexin Chen2Yulin Shao3https://orcid.org/0000-0002-6300-3175Soung Chang Liew4https://orcid.org/0000-0001-7055-6483Department of Information Engineering, The Chinese University of Hong Kong, Hong Kong, SAR, ChinaDepartment of Information Engineering, The Chinese University of Hong Kong, Hong Kong, SAR, ChinaDepartment of Computer Science Engineering, The Chinese University of Hong Kong, Hong Kong, SAR, ChinaDepartment of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, SAR, ChinaDepartment of Information Engineering, The Chinese University of Hong Kong, Hong Kong, SAR, ChinaSemantic communication is a promising technology for next-generation wireless networks. However, the out-of-distribution (OOD) problem, where a pre-trained machine learning (ML) model is applied to unseen tasks that are outside the distribution of its training data, may compromise the integrity of semantic compression. This paper explores the use of multi-modal large language models (MLLMs) to address the OOD issue in image semantic communication. We propose a novel “Plan A - Plan B” framework that leverages the broad knowledge and strong generalization ability of an MLLM to assist a conventional ML model when the latter encounters an OOD input in the semantic encoding process. Furthermore, we propose a Bayesian optimization scheme that reshapes the probability distribution of the MLLM’s inference process based on the contextual information of the image. The optimization scheme significantly enhances the MLLM’s performance in semantic compression by 1) filtering out irrelevant vocabulary in the original MLLM output; and 2) using contextual similarities between prospective answers of the MLLM and the background information as prior knowledge to modify the MLLM’s probability distribution during inference. Further, at the receiver side of the communication system, we put forth a “generate-criticize” framework that utilizes the cooperation of multiple MLLMs to enhance the reliability of image reconstruction.https://ieeexplore.ieee.org/document/11113346/Semantic communicationmulti-modal large language modelgenerative AIsout-of-distribution problem
spellingShingle Feifan Zhang
Yuyang Du
Kexin Chen
Yulin Shao
Soung Chang Liew
Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models
IEEE Transactions on Machine Learning in Communications and Networking
Semantic communication
multi-modal large language model
generative AIs
out-of-distribution problem
title Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models
title_full Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models
title_fullStr Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models
title_full_unstemmed Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models
title_short Out-of-Distribution in Image Semantic Communication: A Solution With Multimodal Large Language Models
title_sort out of distribution in image semantic communication a solution with multimodal large language models
topic Semantic communication
multi-modal large language model
generative AIs
out-of-distribution problem
url https://ieeexplore.ieee.org/document/11113346/
work_keys_str_mv AT feifanzhang outofdistributioninimagesemanticcommunicationasolutionwithmultimodallargelanguagemodels
AT yuyangdu outofdistributioninimagesemanticcommunicationasolutionwithmultimodallargelanguagemodels
AT kexinchen outofdistributioninimagesemanticcommunicationasolutionwithmultimodallargelanguagemodels
AT yulinshao outofdistributioninimagesemanticcommunicationasolutionwithmultimodallargelanguagemodels
AT soungchangliew outofdistributioninimagesemanticcommunicationasolutionwithmultimodallargelanguagemodels