COMPUTER VISION

The effectiveness of MAE pre-pretraining for billion-scale pretraining

October 02, 2023

Abstract

This paper revisits the standard pretrain-then-finetune paradigm used in computer vision for visual recognition tasks. Typically, state-of-the-art foundation models are pretrained using large scale (weakly) supervised datasets with billions of images. We introduce an additional pre-pretraining stage that is simple and uses the self-supervised MAE technique to initialize the model. While MAE has only been shown to scale with the size of models, we find that it scales with the size of the training dataset as well. Thus, our MAE-based pre-pretraining scales with both model and data size making it applicable for training foundation models. Pre-pretraining consistently improves both the model convergence and the downstream transfer performance across a range of model scales (millions to billions of parameters), and dataset sizes (millions to billions of images). We measure the effectiveness of pre-pretraining on 10 different visual recognition tasks spanning image classification, video recognition, object detection, low-shot classification and zero-shot recognition. Our largest model achieves new state-of-the-art results on iNaturalist-18 (91.7%), ImageNet-ReaL (91.1%), 1-shot ImageNet-1k (63.6%), and zero-shot transfer on Food-101 (96.2%). Our study reveals that model initialization plays a significant role, even for web-scale pretraining with billions of images, and our models are available publicly.

Download the Paper

AUTHORS

Written by

Mannat Singh

Quentin Duval

Haoqi Fan

Vaibhav Aggarwal

Aaron Adcock

Piotr Dollar

Christoph Feichtenhofer

Ross Girshick

Rohit Girdhar

Ishan Misra

Publisher

ICCV 2023

Research Topics

Computer Vision

Related Publications

November 11, 2024

COMPUTER VISION

HOI-Swap: Swapping Objects in Videos with Hand-Object Interaction Awareness

Sherry Xue, Romy Luo, Changan Chen, Kristen Grauman

November 11, 2024

October 31, 2024

HUMAN & MACHINE INTELLIGENCE

ROBOTICS

Digitizing Touch with an Artificial Multimodal Fingertip

Mike Lambeta, Tingfan Wu, Ali Sengül, Victoria Rose Most, Nolan Black, Kevin Sawyer, Romeo Mercado, Haozhi Qi, Alexander Sohn, Byron Taylor, Norb Tydingco, Gregg Kammerer, Dave Stroud, Jake Khatha, Kurt Jenkins, Kyle Most, Neal Stein, Ricardo Chavira, Thomas Craven-Bartle, Eric Sanchez, Yitian Ding, Jitendra Malik, Roberto Calandra

October 31, 2024

October 16, 2024

SPEECH & AUDIO

COMPUTER VISION

Movie Gen: A Cast of Media Foundation Models

Movie Gen Team

October 16, 2024

September 10, 2024

COMPUTER VISION

Video Editing via Factorized Diffusion Distillation

Uriel Singer, Amit Zohar, Yuval Kirstain, Shelly Sheynin, Adam Polyak, Devi Parikh, Yaniv Taigman

September 10, 2024

Help Us Pioneer The Future of AI

We share our open source frameworks, tools, libraries, and models for everything from research exploration to large-scale production deployment.