Summary of Masked Autoencoders Are Scalable Learners Of Cellular Morphology, by Oren Kraus et al.
Masked Autoencoders are Scalable Learners of Cellular Morphology
by Oren Kraus, Kian Kenyon-Dean, Saber Saberian, Maryam Fallah, Peter McLean, Jess Leung, Vasudev Sharma, Ayla Khan, Jia Balakrishnan, Safiye Celik, Maciej Sypetkowski, Chi Vicky Cheng, Kristen Morse, Maureen Makes, Ben Mabey, Berton Earnshaw
First submitted to arxiv on: 27 Sep 2023
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research paper explores the use of self-supervised deep learning approaches to infer biological relationships from cellular phenotypes in high-content microscopy screens. The study builds upon prior results showing that deep vision models can capture biological signal better than hand-crafted features, and aims to scale these methods for larger models and datasets. The authors find that both CNN- and ViT-based masked autoencoders outperform weakly supervised baselines, with a ViT-L/8 model trained on over 3.5-billion unique crops sampled from 93-million microscopy images achieving relative improvements of up to 28% compared to the best weakly supervised baseline. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us better understand how cells work and what they do. It uses super powerful computers to look at tiny pictures of cells and figure out which ones are related. The researchers found that by using special computer programs, they can get even more accurate results than before. This is important because it will help scientists learn more about how cells work together and might even lead to new discoveries. |
Keywords
* Artificial intelligence * Cnn * Deep learning * Self supervised * Supervised * Vit
