MultiMAE: Multi-modal Multi-task Masked Autoencoders

4 Apr 2022  ·  Roman Bachmann, David Mizrahi, Andrei Atanov, Amir Zamir ·

We propose a pre-training strategy called Multi-modal Multi-task Masked Autoencoders (MultiMAE). It differs from standard Masked Autoencoding in two key aspects: I) it can optionally accept additional modalities of information in the input besides the RGB image (hence "multi-modal"), and II) its training objective accordingly includes predicting multiple outputs besides the RGB image (hence "multi-task"). We make use of masking (across image patches and input modalities) to make training MultiMAE tractable as well as to ensure cross-modality predictive coding is indeed learned by the network. We show this pre-training strategy leads to a flexible, simple, and efficient framework with improved transfer results to downstream tasks. In particular, the same exact pre-trained network can be flexibly used when additional information besides RGB images is available or when no information other than RGB is available - in all configurations yielding competitive to or significantly better results than the baselines. To avoid needing training datasets with multiple modalities and tasks, we train MultiMAE entirely using pseudo labeling, which makes the framework widely applicable to any RGB dataset. The experiments are performed on multiple transfer tasks (image classification, semantic segmentation, depth estimation) and datasets (ImageNet, ADE20K, Taskonomy, Hypersim, NYUv2). The results show an intriguingly impressive capability by the model in cross-modal/task predictive coding and transfer.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Semantic Segmentation ADE20K MultiMAE (ViT-B) Validation mIoU 46.2 # 174
Semantic Segmentation ADE20K val MultiMAE (ViT-B) mIoU 46.2 # 69
Semantic Segmentation Hypersim MultiMAE (ViT-B) mIoU 37.0 # 2
Semantic Segmentation Hypersim MAE (ViT-B) mIoU 36.5 # 3
Semantic Segmentation Hypersim DINO (ViT-B) mIoU 32.5 # 4
Semantic Segmentation Hypersim MoCo-v3 (ViT-B) mIoU 31.7 # 5
Semantic Segmentation NYU Depth v2 MultiMAE (ViT-B) Mean IoU 56.0% # 17

Methods


No methods listed for this paper. Add relevant methods here