ColorMAE: Exploring data-independent masking strategies in Masked AutoEncoders

Carlos Hinojosa, Shuming Liu, Bernard Ghanem
PDF Supplementary Project Poster DOI

Abstract

We explore data-independent masking strategies for Masked AutoEncoders (MAE), showing that carefully designed structured noise masks can match or improve upon learned/data-dependent masking, while being simpler and more efficient. See our Project Page!

Type
Conference paper
Publication
In 2024 European Conference on Computer Vision
Self-Supervised Learning Masked Autoencoders Computer Vision
Avatar
Carlos Hinojosa
AI Researcher, KAUST

I’m a computer scientist working in computer vision, machine learning, and AI safety. My research focuses on developing accurate, reliable, and efficient vision and AI systems.