LapSurg-230K: Large-Scale Surgical Video Action Recognition Dataset

LapSurg-230K is a large-scale, high-quality surgical video action recognition dataset. It contains 230,000 video clips covering 9 surgery types and 11 fundamental surgical actions. The dataset was authored by Zi Ye and last updated on May 11, 2026.

Use Cases

• Training action recognition models based on labeled surgical video clips.
• Benchmarking surgical AI systems based on the 11 fundamental action classes.
• Developing computer-assisted surgery tools based on the high-quality video data.

Strengths

• Dataset size is 11.9 GB, indicating a substantial collection of video data.
• Covers 9 distinct surgery types, providing diversity in surgical procedures.
• Annotated for 11 fundamental surgical actions, enabling specific model training.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for specific tasks.

Provenance

Source

Zi Ye via figshare

Freshness

Last updated 2026-05-11 10:51:13; freshness should be verified.