Visualizing the loss landscape of Self-supervised Vision Transformer

• URL: http://arxiv.org/abs/2405.18042v1
• Date: Tue, 28 May 2024 10:54:26 GMT
• Title: Visualizing the loss landscape of Self-supervised Vision Transformer
• Authors: Youngwan Lee, Jeffrey Ryan Willette, Jonghee Kim, Sung Ju Hwang,
• Abstract summary: The Masked autoencoder (MAE) has drawn attention as a representative self-supervised approach for masked image modeling with vision transformers. We visualize the loss landscapes of the self-supervised vision transformer by both MAE and RC-MAE and compare them with the supervised ViT (Sup-ViT) To the best of our knowledge, this work is the first to investigate the self-supervised ViT through the lens of the loss landscape.
• Score: 53.84372035496475
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: The Masked autoencoder (MAE) has drawn attention as a representative self-supervised approach for masked image modeling with vision transformers. However, even though MAE shows better generalization capability than fully supervised training from scratch, the reason why has not been explored. In another line of work, the Reconstruction Consistent Masked Auto Encoder (RC-MAE), has been proposed which adopts a self-distillation scheme in the form of an exponential moving average (EMA) teacher into MAE, and it has been shown that the EMA-teacher performs a conditional gradient correction during optimization. To further investigate the reason for better generalization of the self-supervised ViT when trained by MAE (MAE-ViT) and the effect of the gradient correction of RC-MAE from the perspective of optimization, we visualize the loss landscapes of the self-supervised vision transformer by both MAE and RC-MAE and compare them with the supervised ViT (Sup-ViT). Unlike previous loss landscape visualizations of neural networks based on classification task loss, we visualize the loss landscape of ViT by computing pre-training task loss. Through the lens of loss landscapes, we find two interesting observations: (1) MAE-ViT has a smoother and wider overall loss curvature than Sup-ViT. (2) The EMA-teacher allows MAE to widen the region of convexity in both pretraining and linear probing, leading to quicker convergence. To the best of our knowledge, this work is the first to investigate the self-supervised ViT through the lens of the loss landscape.

Related papers

• How Effective is Pre-training of Large Masked Autoencoders for Downstream Earth Observation Tasks? [9.515532265294187]
  Self-supervised pre-training has proven highly effective for many computer vision tasks. It remains unclear under which conditions pre-trained models offer significant advantages over training from scratch.
  arXiv  Detail & Related papers  (2024-09-27T08:15:14Z)
• Attention-Guided Masked Autoencoders For Learning Image Representations [16.257915216763692]
  Masked autoencoders (MAEs) have established themselves as a powerful method for unsupervised pre-training for computer vision tasks. We propose to inform the reconstruction process through an attention-guided loss function. Our evaluations show that our pre-trained models learn better latent representations than the vanilla MAE.
  arXiv  Detail & Related papers  (2024-02-23T08:11:25Z)
• Cross-Scale MAE: A Tale of Multi-Scale Exploitation in Remote Sensing [5.325585142755542]
  We present Cross-Scale MAE, a self-supervised model built upon the Masked Auto-Encoder (MAE).During pre-training, Cross-Scale MAE employs scale augmentation techniques and enforces cross-scale constraints through both contrastive and generative losses. Experimental evaluations demonstrate that Cross-Scale MAE exhibits superior performance compared to standard MAE and other state-of-the-art remote sensing MAE methods.
  arXiv  Detail & Related papers  (2024-01-29T03:06:19Z)
• Understanding Masked Autoencoders From a Local Contrastive Perspective [80.57196495601826]
  Masked AutoEncoder (MAE) has revolutionized the field of self-supervised learning with its simple yet effective masking and reconstruction strategies. We introduce a new empirical framework, called Local Contrastive MAE, to analyze both reconstructive and contrastive aspects of MAE.
  arXiv  Detail & Related papers  (2023-10-03T12:08:15Z)
• Exploring The Role of Mean Teachers in Self-supervised Masked Auto-Encoders [64.03000385267339]
  Masked image modeling (MIM) has become a popular strategy for self-supervised learning(SSL) of visual representations with Vision Transformers. We present a simple SSL method, the Reconstruction-Consistent Masked Auto-Encoder (RC-MAE) by adding an EMA teacher to MAE. RC-MAE converges faster and requires less memory usage than state-of-the-art self-distillation methods during pre-training.
  arXiv  Detail & Related papers  (2022-10-05T08:08:55Z)
• HiViT: Hierarchical Vision Transformer Meets Masked Image Modeling [126.89573619301953]
  We propose a new design of hierarchical vision transformers named HiViT (short for Hierarchical ViT) HiViT enjoys both high efficiency and good performance in MIM. In running MAE on ImageNet-1K, HiViT-B reports a +0.6% accuracy gain over ViT-B and a 1.9$times$ speed-up over Swin-B.
  arXiv  Detail & Related papers  (2022-05-30T09:34:44Z)
• 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)
• 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.