Salvato in:
Dettagli Bibliografici
Autori principali: Moon, JiHwan, Park, Jihoon, Kim, Jungeun, Bae, Jongseong, Jeon, Hyeongwoo, Kim, Ha Young
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2411.17248
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929605806391296
author Moon, JiHwan
Park, Jihoon
Kim, Jungeun
Bae, Jongseong
Jeon, Hyeongwoo
Kim, Ha Young
author_facet Moon, JiHwan
Park, Jihoon
Kim, Jungeun
Bae, Jongseong
Jeon, Hyeongwoo
Kim, Ha Young
contents Sign language translation (SLT) is challenging, as it involves converting sign language videos into natural language. Previous studies have prioritized accuracy over diversity. However, diversity is crucial for handling lexical and syntactic ambiguities in machine translation, suggesting it could similarly benefit SLT. In this work, we propose DiffSLT, a novel gloss-free SLT framework that leverages a diffusion model, enabling diverse translations while preserving sign language semantics. DiffSLT transforms random noise into the target latent representation, conditioned on the visual features of input video. To enhance visual conditioning, we design Guidance Fusion Module, which fully utilizes the multi-level spatiotemporal information of the visual features. We also introduce DiffSLT-P, a DiffSLT variant that conditions on pseudo-glosses and visual features, providing key textual guidance and reducing the modality gap. As a result, DiffSLT and DiffSLT-P significantly improve diversity over previous gloss-free SLT methods and achieve state-of-the-art performance on two SLT datasets, thereby markedly improving translation quality.
format Preprint
id arxiv_https___arxiv_org_abs_2411_17248
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DiffSLT: Enhancing Diversity in Sign Language Translation via Diffusion Model
Moon, JiHwan
Park, Jihoon
Kim, Jungeun
Bae, Jongseong
Jeon, Hyeongwoo
Kim, Ha Young
Computer Vision and Pattern Recognition
Sign language translation (SLT) is challenging, as it involves converting sign language videos into natural language. Previous studies have prioritized accuracy over diversity. However, diversity is crucial for handling lexical and syntactic ambiguities in machine translation, suggesting it could similarly benefit SLT. In this work, we propose DiffSLT, a novel gloss-free SLT framework that leverages a diffusion model, enabling diverse translations while preserving sign language semantics. DiffSLT transforms random noise into the target latent representation, conditioned on the visual features of input video. To enhance visual conditioning, we design Guidance Fusion Module, which fully utilizes the multi-level spatiotemporal information of the visual features. We also introduce DiffSLT-P, a DiffSLT variant that conditions on pseudo-glosses and visual features, providing key textual guidance and reducing the modality gap. As a result, DiffSLT and DiffSLT-P significantly improve diversity over previous gloss-free SLT methods and achieve state-of-the-art performance on two SLT datasets, thereby markedly improving translation quality.
title DiffSLT: Enhancing Diversity in Sign Language Translation via Diffusion Model
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2411.17248