mlsanityData to 7 Oct 2026

Making AI Drafts Count: A Quality Threshold in Audio Description Workflows

arXiv 2605.05348 ↗First posted 6 May 2026cs.HC cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Audio description (AD) narrates visual elements in video for blind and low-vision audiences. Recent work has shown that giving novice describers an AI-generated draft to start from helps produce higher-quality AD and lowers the barrier to entry. What remains an open question is how draft quality shapes the editing process. We investigate this through GenAD, an AD generation pipeline that incorporates accessibility guidelines and contextual video information, and RefineAD, an editing interface for human revisions. Human-AI contributions are measured across text, timing, and delivery. In a within-subjects study, we compared authoring from scratch against editing AI drafts of varying quality. GenAD drafts cut completion time by more than half and significantly reduced cognitive load. In contrast, baseline drafts generated from simple, unguided prompts offered only modest benefits, pointing to a minimum quality threshold for effectiveness. Qualitative findings suggest this threshold is content-dependent; as visual complexity increases, so does the quality needed from AI drafts. We propose this as a design principle: effective AI assistance should clear a quality threshold suited to the target content, rather than simply be present.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 19 of its 51 references that are in the map. First posted on arXiv 6 May 2026.

Builds on

19 of 51 references.

References · 19

  1. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,131
  2. 2024M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillationcited by 1,050
  3. 2023Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formattingcited by 778
  4. 2024ShareGPT4Video: Improving Video Understanding and Generation with Better Captionscited by 386
  5. 2023MM-VID: Advancing Video Understanding with GPT-4V(ision)cited by 82
  6. 2024A Design Space for Intelligent and Interactive Writing Assistantscited by 76
  7. 2024VidHalluc: Evaluating Temporal Hallucinations in Multimodal Large Language Models for Video Understandingcited by 65
  8. 2024AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuancescited by 59
  9. 2024Shaping Human-AI Collaboration: Varied Scaffolding Levels in Co-writing with Language Modelscited by 51
  10. 2024AutoAD-Zero: A Training-Free Framework for Zero-Shot Audio Descriptioncited by 21
  11. 2020Rescribe: Authoring and Automatically Editing Audio Descriptionscited by 20
  12. 2023AVscript: Accessible Video Editing with Audio-Visual Scriptscited by 19
7 more
  1. 2023The Role of Inclusion, Control, and Ownership in Workplace AI-Mediated Communicationcited by 19
  2. 2024Making Short-Form Videos Accessible with Hierarchical Video Summariescited by 14
  3. 2024LLM-AD: Large Language Model based Audio Description Systemcited by 14
  4. 2025VideoA11y: Method and Dataset for Accessible Video Descriptioncited by 8
  5. 2022CrossA11y: Identifying Video Accessibility Issues via Cross-modal Groundingcited by 6
  6. 2024Describe Now: User-Driven Audio Description for Blind and Low Vision Individualscited by 4
  7. 2026ADCanvas: Accessible and Conversational Audio Description Authoring for Blind and Low Vision Creatorscited by 3

Built on by

Not cited yet.