AI BENCHMARK PROFILE
StrAD
StrAD is a benchmark for long-form audio description generation on full-length videos across diverse genres, reformulating the task as streaming dense video captioning without ground-truth timestamps.
- Released
- 2026-08-12
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Addresses the lack of evaluation for full-video AD generation, providing a measurable benchmark that includes both fine-tuned and zero-shot approaches, which is essential for scaling accessibility.
Motivation
Visual content is the dominant medium of communication, yet without audio descriptions (ADs), it remains inaccessible to blind and low-vision people.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.