{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/zero-avsr-zero-shot-audio-visual-speech","title":"Zero-AVSR: Zero-Shot Audio-Visual Speech Recognition with LLMs by Learning Language-Agnostic Speech Representations","arxiv_id":"2503.06273","date":"2025-03-08","proceeding":null,"authors":["Jeong Hun Yeo","Minsu Kim","Chae Won Kim","Stavros Petridis","Yong Man Ro"],"abstract":"We explore a novel zero-shot Audio-Visual Speech Recognition (AVSR) framework, dubbed Zero-AVSR, which enables speech recognition in target languages without requiring any audio-visual speech data in those languages. Specifically, we introduce the Audio-Visual Speech Romanizer (AV-Romanizer), which learns language-agnostic speech representations by predicting Roman text. Then, by leveraging the strong multilingual modeling capabilities of Large Language Models (LLMs), we propose converting the predicted Roman text into language-specific graphemes, forming the proposed Cascaded Zero-AVSR. Taking it a step further, we explore a unified Zero-AVSR approach by directly integrating the audio-visual speech representations encoded by the AV-Romanizer into the LLM. This is achieved through finetuning the adapter and the LLM using our proposed multi-task learning scheme. To capture the wide spectrum of phonetic and linguistic diversity, we also introduce a Multilingual Audio-Visual Romanized Corpus (MARC) consisting of 2,916 hours of audio-visual speech data across 82 languages, along with transcriptions in both language-specific graphemes and Roman text. Extensive analysis and experiments confirm that the proposed Zero-AVSR framework has the potential to expand language support beyond the languages seen during the training of the AV-Romanizer.","url_abs":"https://arxiv.org/abs/2503.06273v1","url_pdf":"https://arxiv.org/pdf/2503.06273v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"zero-avsr-zero-shot-audio-visual-speech","repo_url":"https://github.com/JeongHun0716/zero-avsr","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"audio-visual-speech-recognition","task_name":"Audio-Visual Speech Recognition"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"},{"task_slug":"speech-recognition","task_name":"Speech Recognition"},{"task_slug":"visual-speech-recognition","task_name":"Visual Speech Recognition"},{"task_slug":"speech-recognition-1","task_name":"speech-recognition"}],"methods":[{"method_slug":"adapter","method_name":"Adapter"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/audio-visual-speech-recognition-on-lrs3-ted","task":"Audio-Visual Speech Recognition","dataset":"LRS3-TED","model":"Zero-AVSR","rank_in_archive_order":8,"of":12,"metrics":{"Word Error Rate (WER)":"1.5"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2503.06273","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}