Spotlight
in
Workshop: Generative AI and Biology (GenBio@NeurIPS2023)
Masked autoencoders are scalable learners of cellular morphology
Oren Kraus · Kian Kenyon-Dean · Saber Saberian · Maryam Fallah · Peter McLean · Jess Leung · Vasudev Sharma · Ayla Khan · Jia Balakrishnan · Safiye Celik · Maciej Sypetkowski · Chi Cheng · Kristen Morse · Maureen Makes · Ben Mabey · Berton Earnshaw
Keywords: [ Vision transformer ] [ Foundation Model ] [ Computer Vision ] [ Masked Autoencoder ] [ microscopy ] [ high content screening ] [ CRISPR ]
Sat 16 Dec 6:15 a.m. PST — 3:30 p.m. PST
Inferring biological relationships from cellular phenotypes in high-content microscopy screens provides significant opportunity and challenge in biological research. Prior results have shown that deep vision models can capture biological signal better than hand-crafted features. This work explores how weakly supervised and self-supervised deep learning approaches scale when training larger models on larger datasets. Our results show that both CNN- and ViT-based masked autoencoders significantly outperform weakly supervised models. At the high-end of our scale, a ViT-L/8 trained on over 3.5-billion unique crops sampled from 95-million microscopy images achieves relative improvements as high as 28% over our best weakly supervised models at inferring known biological relationships curated from public databases.