Andha-Dhun: A First Look at Audio Descriptions in Hindi
Not on the map. Why?
Abstract (arXiv)
Audio Descriptions (ADs) narrate visual content for Blind and Low Vision (BLV) audiences during gaps in audiovisual media. There is growing momentum around ADs in movies and TV shows, and with mandates from India's Central Board of Film Certification (CBFC), there is a need to expand ADs beyond English. Yet, there is no work that generates ADs for any Indian language. To address this gap, we present the first systematic study of ADs in Hindi, contributing to aspects such as data, generation, and evaluation. We introduce Andha-Dhun, the first dataset of human-authored Hindi ADs collected from 8 full-length movies. We explore two approaches for generating ADs in Hindi: (i) directly from English dense video descriptions, and (ii) translating English ADs into Hindi. We evaluate these approaches using perplexity and LLM-as-a-judge metrics to assess fluency and quality respectively. We also analyze movies that have both English and Hindi human-authored ADs and find that naive translation introduces artifacts and narrows diversity compared to original Hindi ADs. Direct machine translation fails to adapt cultural references, while human-translated ADs do better but still fall short. Our findings emphasize that the purpose of Hindi ADs is accessibility for Indian BLV audiences, and that this requires adapting content for the audience more than strict fidelity to the source.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 25 of its 55 references that are in the map. First posted on arXiv 7 Jul 2026.
Builds on
25 of 55 references.
References · 25
- 2024The Llama 3 Herd of Modelscited by 14,715
- 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
- 2022No Language Left Behind: Scaling Human-Centered Machine Translationcited by 1,312
- 2024From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judgecited by 494
- 2023WhisperX: Time-Accurate Speech Transcription of Long-Form Audiocited by 326
- 2023MADLAD-400: A Multilingual And Document-Level Large Audited Datasetcited by 181
- 2023IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languagescited by 152
- 2019Unifying Human and Statistical Evaluation for Natural Language Generationcited by 127
- 2021MAD: A Scalable Dataset for Language Grounding in Videos from Movie Audio Descriptionscited by 108
- 2021Survey of Low-Resource Machine Translationcited by 92
- 2023AutoAD: Movie Description in Contextcited by 50
- 2023AutoAD II: The Sequel -- Who, When, and What in Movie Audio Descriptioncited by 49
13 more
- 2023MM-Narrator: Narrating Long-form Videos with Multimodal In-Context Learningcited by 46
- 2024AutoAD III: The Prequel -- Back to the Pixelscited by 35
- 2024Adapting Multilingual LLMs to Low-Resource Languages using Continued Pre-training and Synthetic Corpuscited by 24
- 2024AutoAD-Zero: A Training-Free Framework for Zero-Shot Audio Descriptioncited by 23
- 2019Judge the Judges: A Large-Scale Evaluation Study of Neural Language Models for Online Review Generationcited by 17
- 2024Learning Video Context as Interleaved Multimodal Sequencescited by 16
- 2024LLM-AD: Large Language Model based Audio Description Systemcited by 15
- 2024Contextual AD Narration with Interleaved Multimodal Sequencecited by 14
- 2025Shot-by-Shot: Film-Grammar-Aware Training-Free Audio Description Generationcited by 13
- 2024DistinctAD: Distinctive Audio Description Generation in Contextscited by 12
- 2025What You See is What You Ask: Evaluating Audio Descriptionscited by 8
- 2024Audio Description Generation in the Era of LLMs and VLMs: A Review of Transferable Generative AI Technologiescited by 5
- 2025More than a Moment: Towards Coherent Sequences of Audio Descriptionscited by 5
Built on by
Not cited yet.