On the Effectiveness of Integration Methods for Multimodal Dialogue Response Retrieval

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jang, Seongbo, Lee, Seonghyeon, Lee, Dongha, Yu, Hwanjo
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913083970027520
author Jang, Seongbo
Lee, Seonghyeon
Lee, Dongha
Yu, Hwanjo
author_facet Jang, Seongbo
Lee, Seonghyeon
Lee, Dongha
Yu, Hwanjo
contents Multimodal chatbots have become one of the major topics for dialogue systems in both research community and industry. Recently, researchers have shed light on the multimodality of responses as well as dialogue contexts. This work explores how a dialogue system can output responses in various modalities such as text and image. To this end, we first formulate a multimodal dialogue response retrieval task for retrieval-based systems as the combination of three subtasks. We then propose three integration methods based on a two-step approach and an end-to-end approach, and compare the merits and demerits of each method. Experimental results on two datasets demonstrate that the end-to-end approach achieves comparable performance without an intermediate step in the two-step approach. In addition, a parameter sharing strategy not only reduces the number of parameters but also boosts performance by transferring knowledge across the subtasks and the modalities.
format Preprint
id arxiv_https___arxiv_org_abs_2506_11499
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the Effectiveness of Integration Methods for Multimodal Dialogue Response Retrieval
Jang, Seongbo
Lee, Seonghyeon
Lee, Dongha
Yu, Hwanjo
Computation and Language
Multimodal chatbots have become one of the major topics for dialogue systems in both research community and industry. Recently, researchers have shed light on the multimodality of responses as well as dialogue contexts. This work explores how a dialogue system can output responses in various modalities such as text and image. To this end, we first formulate a multimodal dialogue response retrieval task for retrieval-based systems as the combination of three subtasks. We then propose three integration methods based on a two-step approach and an end-to-end approach, and compare the merits and demerits of each method. Experimental results on two datasets demonstrate that the end-to-end approach achieves comparable performance without an intermediate step in the two-step approach. In addition, a parameter sharing strategy not only reduces the number of parameters but also boosts performance by transferring knowledge across the subtasks and the modalities.
title On the Effectiveness of Integration Methods for Multimodal Dialogue Response Retrieval
topic Computation and Language
url https://arxiv.org/abs/2506.11499