Learning Self-Regularized Adversarial Views for Self-Supervised Vision Transformers

• URL: http://arxiv.org/abs/2210.08458v1
• Date: Sun, 16 Oct 2022 06:20:44 GMT
• Title: Learning Self-Regularized Adversarial Views for Self-Supervised Vision Transformers
• Authors: Tao Tang, Changlin Li, Guangrun Wang, Kaicheng Yu, Xiaojun Chang, Xiaodan Liang
• Abstract summary: We propose a self-regularized AutoAugment method to learn views for self-supervised vision transformers. First, we reduce the search cost of AutoView to nearly zero by learning views and network parameters simultaneously. We also present a curated augmentation policy search space for self-supervised learning.
• Score: 105.89564687747134
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Automatic data augmentation (AutoAugment) strategies are indispensable in supervised data-efficient training protocols of vision transformers, and have led to state-of-the-art results in supervised learning. Despite the success, its development and application on self-supervised vision transformers have been hindered by several barriers, including the high search cost, the lack of supervision, and the unsuitable search space. In this work, we propose AutoView, a self-regularized adversarial AutoAugment method, to learn views for self-supervised vision transformers, by addressing the above barriers. First, we reduce the search cost of AutoView to nearly zero by learning views and network parameters simultaneously in a single forward-backward step, minimizing and maximizing the mutual information among different augmented views, respectively. Then, to avoid information collapse caused by the lack of label supervision, we propose a self-regularized loss term to guarantee the information propagation. Additionally, we present a curated augmentation policy search space for self-supervised learning, by modifying the generally used search space designed for supervised learning. On ImageNet, our AutoView achieves remarkable improvement over RandAug baseline (+10.2% k-NN accuracy), and consistently outperforms sota manually tuned view policy by a clear margin (up to +1.3% k-NN accuracy). Extensive experiments show that AutoView pretraining also benefits downstream tasks (+1.2% mAcc on ADE20K Semantic Segmentation and +2.8% mAP on revisited Oxford Image Retrieval benchmark) and improves model robustness (+2.3% Top-1 Acc on ImageNet-A and +1.0% AUPR on ImageNet-O). Code and models will be available at https://github.com/Trent-tangtao/AutoView.

Related papers

• Federated Self-Supervised Learning of Monocular Depth Estimators for Autonomous Vehicles [0.0]
  FedSCDepth is a novel method that combines federated learning and deep self-supervision to enable the learning of monocular depth estimators. Our proposed method achieves near state-of-the-art performance, with a test loss below 0.13 and requiring, on average, only 1.5k training steps.
  arXiv  Detail & Related papers  (2023-10-07T14:54:02Z)
• A Novel Driver Distraction Behavior Detection Method Based on Self-supervised Learning with Masked Image Modeling [5.1680226874942985]
  Driver distraction causes a significant number of traffic accidents every year, resulting in economic losses and casualties. Driver distraction detection primarily relies on traditional convolutional neural networks (CNN) and supervised learning methods. This paper proposes a new self-supervised learning method based on masked image modeling for driver distraction behavior detection.
  arXiv  Detail & Related papers  (2023-06-01T10:53:32Z)
• M$^2$DAR: Multi-View Multi-Scale Driver Action Recognition with Vision Transformer [5.082919518353888]
  We present a multi-view, multi-scale framework for naturalistic driving action recognition and localization in untrimmed videos. Our system features a weight-sharing, multi-scale Transformer-based action recognition network that learns robust hierarchical representations.
  arXiv  Detail & Related papers  (2023-05-13T02:38:15Z)
• Self-Supervised Representation Learning from Temporal Ordering of Automated Driving Sequences [49.91741677556553]
  We propose TempO, a temporal ordering pretext task for pre-training region-level feature representations for perception tasks. We embed each frame by an unordered set of proposal feature vectors, a representation that is natural for object detection or tracking systems. Extensive evaluations on the BDD100K, nuImages, and MOT17 datasets show that our TempO pre-training approach outperforms single-frame self-supervised learning methods.
  arXiv  Detail & Related papers  (2023-02-17T18:18:27Z)
• AdaViT: Adaptive Tokens for Efficient Vision Transformer [91.88404546243113]
  We introduce AdaViT, a method that adaptively adjusts the inference cost of vision transformer (ViT) for images of different complexity. AdaViT achieves this by automatically reducing the number of tokens in vision transformers that are processed in the network as inference proceeds.
  arXiv  Detail & Related papers  (2021-12-14T18:56:07Z)
• Efficient Self-supervised Vision Transformers for Representation Learning [86.57557009109411]
  We show that multi-stage architectures with sparse self-attentions can significantly reduce modeling complexity. We propose a new pre-training task of region matching which allows the model to capture fine-grained region dependencies. Our results show that combining the two techniques, EsViT achieves 81.3% top-1 on the ImageNet linear probe evaluation.
  arXiv  Detail & Related papers  (2021-06-17T19:57:33Z)
• Fine-Grained Vehicle Perception via 3D Part-Guided Visual Data Augmentation [77.60050239225086]
  We propose an effective training data generation process by fitting a 3D car model with dynamic parts to vehicles in real images. Our approach is fully automatic without any human interaction. We present a multi-task network for VUS parsing and a multi-stream network for VHI parsing.
  arXiv  Detail & Related papers  (2020-12-15T03:03:38Z)
• Auto-Rectify Network for Unsupervised Indoor Depth Estimation [119.82412041164372]
  We establish that the complex ego-motions exhibited in handheld settings are a critical obstacle for learning depth. We propose a data pre-processing method that rectifies training images by removing their relative rotations for effective learning. Our results outperform the previous unsupervised SOTA method by a large margin on the challenging NYUv2 dataset.
  arXiv  Detail & Related papers  (2020-06-04T08:59:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.