Codesota · Models · SegVolBAAI (Beijing Academy of AI)1 results · 1 benchmarks
Model card

SegVol.

BAAI (Beijing Academy of AI)open-source90M paramsSAM-based volumetric segmentation

SAM-based universal volumetric segmentation model. Supports semantic and spatial prompts. Trained on 90K unlabeled CT scans.

§ 01 · Benchmarks

Every benchmark SegVol has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01Synapse Multi-Organ CTMedical · Medical Image Segmentationmean-dsc83.1%#6/112023-11-22source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 02 · Strengths by area

Where SegVol actually performs.

Medical
1
benchmark
avg rank #6.0
§ 03 · Papers

1 paper with results for SegVol.

  1. 2023-11-22· Medical· 1 result

    SegVol: Universal and Interactive Volumetric Medical Image Segmentation

§ 05 · Sources & freshness

Where these numbers come from.

arxiv
1
result
1 of 1 rows marked verified.