Seeing What You Say: Expressive Image Generation from Speech

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Lee, Jiyoung, Park, Song, Chun, Sanghyuk, Chung, Soo-Whan
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915600033382400
author Lee, Jiyoung
Park, Song
Chun, Sanghyuk
Chung, Soo-Whan
author_facet Lee, Jiyoung
Park, Song
Chun, Sanghyuk
Chung, Soo-Whan
contents This paper proposes VoxStudio, the first unified and end-to-end speech-to-image model that generates expressive images directly from spoken descriptions by jointly aligning linguistic and paralinguistic information. At its core is a speech information bottleneck (SIB) module, which compresses raw speech into compact semantic tokens, preserving prosody and emotional nuance. By operating directly on these tokens, VoxStudio eliminates the need for an additional speech-to-text system, which often ignores the hidden details beyond text, e.g., tone or emotion. We also release VoxEmoset, a large-scale paired emotional speech-image dataset built via an advanced TTS engine to affordably generate richly expressive utterances. Comprehensive experiments on the SpokenCOCO, Flickr8kAudio, and VoxEmoset benchmarks demonstrate the feasibility of our method and highlight key challenges, including emotional consistency and linguistic ambiguity, paving the way for future research.
format Preprint
id arxiv_https___arxiv_org_abs_2511_03423
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Seeing What You Say: Expressive Image Generation from Speech
Lee, Jiyoung
Park, Song
Chun, Sanghyuk
Chung, Soo-Whan
Audio and Speech Processing
Computer Vision and Pattern Recognition
Multimedia
This paper proposes VoxStudio, the first unified and end-to-end speech-to-image model that generates expressive images directly from spoken descriptions by jointly aligning linguistic and paralinguistic information. At its core is a speech information bottleneck (SIB) module, which compresses raw speech into compact semantic tokens, preserving prosody and emotional nuance. By operating directly on these tokens, VoxStudio eliminates the need for an additional speech-to-text system, which often ignores the hidden details beyond text, e.g., tone or emotion. We also release VoxEmoset, a large-scale paired emotional speech-image dataset built via an advanced TTS engine to affordably generate richly expressive utterances. Comprehensive experiments on the SpokenCOCO, Flickr8kAudio, and VoxEmoset benchmarks demonstrate the feasibility of our method and highlight key challenges, including emotional consistency and linguistic ambiguity, paving the way for future research.
title Seeing What You Say: Expressive Image Generation from Speech
topic Audio and Speech Processing
Computer Vision and Pattern Recognition
Multimedia
url https://arxiv.org/abs/2511.03423