BatchBench: Benchmarking AI Robustness to Batch Effects and Domain Shift in Biomedical Images - A JUMP Cell Painting Leaderboard
BatchBench tests how well AI handles biomedical images from experimental batches and data sources it has not seen before.
The leaderboard uses JUMP-CP, one of the largest publicly available biomedical imaging datasets, generated with the Cell Painting assay. Its pronounced batch effects and domain shifts make it a challenging test of robustness across experimental conditions. The authors say significant performance gaps persist despite advances in domain adaptation.
ArXiv · AI/CL/LG's note
The leaderboard uses JUMP-CP, one of the largest publicly available biomedical imaging datasets, generated with the Cell Painting assay. Its pronounced batch effects and domain shifts make it a challenging test of robustness across experimental conditions. The authors say significant performance gaps persist despite advances in domain adaptation.
ArXiv · AI/CL/LG's note
score 4