Saved in:
Bibliographic Details
Main Authors: Guo, Zirun, Jin, Tao, Xu, Wenlong, Lin, Wang, Wu, Yangyang
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.07121
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912225151680512
author Guo, Zirun
Jin, Tao
Xu, Wenlong
Lin, Wang
Wu, Yangyang
author_facet Guo, Zirun
Jin, Tao
Xu, Wenlong
Lin, Wang
Wu, Yangyang
contents Multimodal sentiment analysis (MSA) is an emerging research topic that aims to understand and recognize human sentiment or emotions through multiple modalities. However, in real-world dynamic scenarios, the distribution of target data is always changing and different from the source data used to train the model, which leads to performance degradation. Common adaptation methods usually need source data, which could pose privacy issues or storage overheads. Therefore, test-time adaptation (TTA) methods are introduced to improve the performance of the model at inference time. Existing TTA methods are always based on probabilistic models and unimodal learning, and thus can not be applied to MSA which is often considered as a multimodal regression task. In this paper, we propose two strategies: Contrastive Adaptation and Stable Pseudo-label generation (CASP) for test-time adaptation for multimodal sentiment analysis. The two strategies deal with the distribution shifts for MSA by enforcing consistency and minimizing empirical risk, respectively. Extensive experiments show that CASP brings significant and consistent improvements to the performance of the model across various distribution shift settings and with different backbones, demonstrating its effectiveness and versatility. Our codes are available at https://github.com/zrguo/CASP.
format Preprint
id arxiv_https___arxiv_org_abs_2412_07121
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Bridging the Gap for Test-Time Multimodal Sentiment Analysis
Guo, Zirun
Jin, Tao
Xu, Wenlong
Lin, Wang
Wu, Yangyang
Machine Learning
Computation and Language
Multimodal sentiment analysis (MSA) is an emerging research topic that aims to understand and recognize human sentiment or emotions through multiple modalities. However, in real-world dynamic scenarios, the distribution of target data is always changing and different from the source data used to train the model, which leads to performance degradation. Common adaptation methods usually need source data, which could pose privacy issues or storage overheads. Therefore, test-time adaptation (TTA) methods are introduced to improve the performance of the model at inference time. Existing TTA methods are always based on probabilistic models and unimodal learning, and thus can not be applied to MSA which is often considered as a multimodal regression task. In this paper, we propose two strategies: Contrastive Adaptation and Stable Pseudo-label generation (CASP) for test-time adaptation for multimodal sentiment analysis. The two strategies deal with the distribution shifts for MSA by enforcing consistency and minimizing empirical risk, respectively. Extensive experiments show that CASP brings significant and consistent improvements to the performance of the model across various distribution shift settings and with different backbones, demonstrating its effectiveness and versatility. Our codes are available at https://github.com/zrguo/CASP.
title Bridging the Gap for Test-Time Multimodal Sentiment Analysis
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2412.07121