Are Caption Metrics Broken? Latency, Deaf and Hard of Hearing User Ratings, and Bias across Technologies
研究通过大规模在线调查评估电视字幕质量，对比不同技术下的字幕延迟对聋人和重听观众的影响，发现现有度量标准并不中立。
0 citationsRead paper
Selected work
Representative Papers
Bigger than the EAR BOX: A Theory-Grounded Review of XR Accessibility Research for Deaf and Hard of Hearing Communities
本文通过构建理论框架分析了53项XR技术在聋哑及重听群体中的应用研究，指出当前研究忽视了该群体多样性，并提出改进方向。
0 citationsRead paper
Seeing the Voice, Preserving the Self: A Participatory Design Approach to Deaf-Centric Text-to-Speech
本文采用参与式设计方法，针对聋哑人及重听者在使用文本转语音技术时遇到的问题，如情感表达和文化身份尊重等，探索了更符合他们需求的解决方案。
0 citationsRead paper
DiscoSign: Discourse-Aware Text to Sign Language Gloss Translation
本文提出DiscoSign，一种基于大语言模型的文本到手语词汇翻译方法，解决了空间共指、问答从句和概念-词汇一致性等话语现象问题。
0 citationsRead paper
Barriers to Employment: The Deaf Multimedia Authoring Tax
This study identifies systemic accessibility barriers faced by Deaf and hard-of-hearing (DHH) individuals when creating multimedia content—such as portfolios, instructional videos, and presentation recordings—imposing disproportionate time and resource burdens across hiring, onboarding, and job performance: a phenomenon termed the “Deaf content creation tax.” Through empirical case analysis, participatory observation, and human factors–informed evaluation of inclusive workflows, the study formally defines this concept for the first time, positioning production-side inaccessibility as a critical mechanism underlying employment inequity. It identifies seven core barrier categories and proposes twelve actionable, evidence-based design guidelines. These guidelines offer standardized adaptation pathways for educational institutions, employers, and content platforms, substantially reducing content production costs for DHH creators. The work bridges a critical theoretical and practical gap in accessibility research by centering content *production*—not just consumption—as a domain requiring systematic intervention.
0 citationsRead paper
Recent publications
Latest Papers
Are Caption Metrics Broken? Latency, Deaf and Hard of Hearing User Ratings, and Bias across Technologies
研究通过大规模在线调查评估电视字幕质量，对比不同技术下的字幕延迟对聋人和重听观众的影响，发现现有度量标准并不中立。
0 citationsRead paper
Bigger than the EAR BOX: A Theory-Grounded Review of XR Accessibility Research for Deaf and Hard of Hearing Communities
本文通过构建理论框架分析了53项XR技术在聋哑及重听群体中的应用研究，指出当前研究忽视了该群体多样性，并提出改进方向。
0 citationsRead paper
Seeing the Voice, Preserving the Self: A Participatory Design Approach to Deaf-Centric Text-to-Speech
本文采用参与式设计方法，针对聋哑人及重听者在使用文本转语音技术时遇到的问题，如情感表达和文化身份尊重等，探索了更符合他们需求的解决方案。
0 citationsRead paper
DiscoSign: Discourse-Aware Text to Sign Language Gloss Translation
本文提出DiscoSign，一种基于大语言模型的文本到手语词汇翻译方法，解决了空间共指、问答从句和概念-词汇一致性等话语现象问题。
0 citationsRead paper
Barriers to Employment: The Deaf Multimedia Authoring Tax
This study identifies systemic accessibility barriers faced by Deaf and hard-of-hearing (DHH) individuals when creating multimedia content—such as portfolios, instructional videos, and presentation recordings—imposing disproportionate time and resource burdens across hiring, onboarding, and job performance: a phenomenon termed the “Deaf content creation tax.” Through empirical case analysis, participatory observation, and human factors–informed evaluation of inclusive workflows, the study formally defines this concept for the first time, positioning production-side inaccessibility as a critical mechanism underlying employment inequity. It identifies seven core barrier categories and proposes twelve actionable, evidence-based design guidelines. These guidelines offer standardized adaptation pathways for educational institutions, employers, and content platforms, substantially reducing content production costs for DHH creators. The work bridges a critical theoretical and practical gap in accessibility research by centering content *production*—not just consumption—as a domain requiring systematic intervention.
0 citationsRead paper