MoPro: Webly Supervised Learning with Momentum PrototypesDownload PDF

28 Sept 2020, 15:47 (modified: 10 Feb 2022, 11:46)ICLR 2021 PosterReaders: Everyone
Keywords: webly-supervised learning, weakly-supervised learning, contrastive learning, representation learning
Abstract: We propose a webly-supervised representation learning method that does not suffer from the annotation unscalability of supervised learning, nor the computation unscalability of self-supervised learning. Most existing works on webly-supervised representation learning adopt a vanilla supervised learning method without accounting for the prevalent noise in the training data, whereas most prior methods in learning with label noise are less effective for real-world large-scale noisy data. We propose momentum prototypes (MoPro), a simple contrastive learning method that achieves online label noise correction, out-of-distribution sample removal, and representation learning. MoPro achieves state-of-the-art performance on WebVision, a weakly-labeled noisy dataset. MoPro also shows superior performance when the pretrained model is transferred to down-stream image classification and detection tasks. It outperforms the ImageNet supervised pretrained model by +10.5 on 1-shot classification on VOC, and outperforms the best self-supervised pretrained model by +17.3 when finetuned on 1% of ImageNet labeled samples. Furthermore, MoPro is more robust to distribution shifts. Code and pretrained models are available at https://github.com/salesforce/MoPro.
One-sentence Summary: MoPro is a new webly-supervised learning framework which advances representation learning using freely-available Web images.
Supplementary Material: zip
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Code: [![github](/images/github_icon.svg) salesforce/MoPro](https://github.com/salesforce/MoPro)
Data: [COCO](https://paperswithcode.com/dataset/coco), [ImageNet](https://paperswithcode.com/dataset/imagenet), [ImageNet-A](https://paperswithcode.com/dataset/imagenet-a), [ImageNet-R](https://paperswithcode.com/dataset/imagenet-r), [Places205](https://paperswithcode.com/dataset/places205), [WebVision](https://paperswithcode.com/dataset/webvision-database)
9 Replies

Loading