Benchmark RadarSEAM-Bench
SEAM-Bench is a double-blind continuity storyboarding benchmark for evaluating visual continuity in short-drama generation. It assesses cross-episode continuity recall and generalizes across six mainstream text models.
- Released
- 2026-08-24
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
The benchmark fills the gap in evaluating visual continuity in large-scale short-drama generation, which is critical for production pipelines. It provides a standardized protocol to compare memory-based approaches and guide improvements in episodic generation.
Motivation
Short-drama generation has grown into a large, industrialized pipeline, and as it scales from isolated shots to the episode level, visual continuity has become a critical bottleneck.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.