Harnessing small projectors and multiple views for efficient vision pretraining

• URL: http://arxiv.org/abs/2312.10725v2
• Date: Fri, 17 Jan 2025 07:01:43 GMT
• Title: Harnessing small projectors and multiple views for efficient vision pretraining
• Authors: Kumar Krishna Agrawal, Arna Ghosh, Shagun Sodhani, Adam Oberman, Blake Richards,
• Abstract summary: We build on recent analytical results to design practical recommendations for competitive and efficient visual representation learning. We show how this idealized loss can be reformulated to a functionally equivalent loss that is more efficient to compute. We empirically verify our findings on CIFAR, STL and Imagenet datasets.
• Score: 11.325655646957186
• License:
• Abstract: Recent progress in self-supervised (SSL) visual representation learning has led to the development of several different proposed frameworks that rely on augmentations of images but use different loss functions. However, there are few theoretically grounded principles to guide practice, so practical implementation of each SSL framework requires several heuristics to achieve competitive performance. In this work, we build on recent analytical results to design practical recommendations for competitive and efficient SSL that are grounded in theory. Specifically, recent theory tells us that existing SSL frameworks are minimizing the same idealized loss, which is to learn features that best match the data similarity kernel defined by the augmentations used. We show how this idealized loss can be reformulated to a functionally equivalent loss that is more efficient to compute. We study the implicit bias of using gradient descent to minimize our reformulated loss function and find that using a stronger orthogonalization constraint with a reduced projector dimensionality should yield good representations. Furthermore, the theory tells us that approximating the reformulated loss should be improved by increasing the number of augmentations, and as such using multiple augmentations should lead to improved convergence. We empirically verify our findings on CIFAR, STL and Imagenet datasets, wherein we demonstrate an improved linear readout performance when training a ResNet-backbone using our theoretically grounded recommendations. Remarkably, we also demonstrate that by leveraging these insights, we can reduce the pretraining dataset size by up to 2$\times$ while maintaining downstream accuracy simply by using more data augmentations. Taken together, our work provides theoretically grounded recommendations that can be used to improve SSL convergence and efficiency.

Related papers

• Dynamic Loss-Based Sample Reweighting for Improved Large Language Model Pretraining [55.262510814326035]
  Existing reweighting strategies primarily focus on group-level data importance. We introduce novel algorithms for dynamic, instance-level data reweighting. Our framework allows us to devise reweighting strategies deprioritizing redundant or uninformative data.
  arXiv  Detail & Related papers  (2025-02-10T17:57:15Z)
• Towards Robust Out-of-Distribution Generalization: Data Augmentation and Neural Architecture Search Approaches [4.577842191730992]
  We study ways toward robust OoD generalization for deep learning. We first propose a novel and effective approach to disentangle the spurious correlation between features that are not essential for recognition. We then study the problem of strengthening neural architecture search in OoD scenarios.
  arXiv  Detail & Related papers  (2024-10-25T20:50:32Z)
• On Improving the Algorithm-, Model-, and Data- Efficiency of Self-Supervised Learning [18.318758111829386]
  We propose an efficient single-branch SSL method based on non-parametric instance discrimination. We also propose a novel self-distillation loss that minimizes the KL divergence between the probability distribution and its square root version.
  arXiv  Detail & Related papers  (2024-04-30T06:39:04Z)
• Random Linear Projections Loss for Hyperplane-Based Optimization in Neural Networks [22.348887008547653]
  This work introduces Random Linear Projections (RLP) loss, a novel approach that enhances training efficiency by leveraging geometric relationships within the data. Our empirical evaluations, conducted across benchmark datasets and synthetic examples, demonstrate that neural networks trained with RLP loss outperform those trained with traditional loss functions.
  arXiv  Detail & Related papers  (2023-11-21T05:22:39Z)
• Gradient constrained sharpness-aware prompt learning for vision-language models [99.74832984957025]
  This paper targets a novel trade-off problem in generalizable prompt learning for vision-language models (VLM) By analyzing the loss landscapes of the state-of-the-art method and vanilla Sharpness-aware Minimization (SAM) based method, we conclude that the trade-off performance correlates to both loss value and loss sharpness. We propose a novel SAM-based method for prompt learning, denoted as Gradient Constrained Sharpness-aware Context Optimization (GCSCoOp)
  arXiv  Detail & Related papers  (2023-09-14T17:13:54Z)
• Towards a Better Theoretical Understanding of Independent Subnetwork Training [56.24689348875711]
  We take a closer theoretical look at Independent Subnetwork Training (IST) IST is a recently proposed and highly effective technique for solving the aforementioned problems. We identify fundamental differences between IST and alternative approaches, such as distributed methods with compressed communication.
  arXiv  Detail & Related papers  (2023-06-28T18:14:22Z)
• Unifying Synergies between Self-supervised Learning and Dynamic Computation [53.66628188936682]
  We present a novel perspective on the interplay between SSL and DC paradigms. We show that it is feasible to simultaneously learn a dense and gated sub-network from scratch in a SSL setting. The co-evolution during pre-training of both dense and gated encoder offers a good accuracy-efficiency trade-off.
  arXiv  Detail & Related papers  (2023-01-22T17:12:58Z)
• Does Self-supervised Learning Really Improve Reinforcement Learning from Pixels? [42.404871049605084]
  We extend the contrastive reinforcement learning framework (e.g., CURL) that jointly optimize SSL and RL losses. Our observations suggest that the existing SSL framework for RL fails to bring meaningful improvement over the baselines. We evaluate the approach in multiple different environments including a real-world robot environment.
  arXiv  Detail & Related papers  (2022-06-10T17:59:30Z)
• Revisiting Consistency Regularization for Semi-Supervised Learning [80.28461584135967]
  We propose an improved consistency regularization framework by a simple yet effective technique, FeatDistLoss. Experimental results show that our model defines a new state of the art for various datasets and settings.
  arXiv  Detail & Related papers  (2021-12-10T20:46:13Z)
• Extrapolation for Large-batch Training in Deep Learning [72.61259487233214]
  We show that a host of variations can be covered in a unified framework that we propose. We prove the convergence of this novel scheme and rigorously evaluate its empirical performance on ResNet, LSTM, and Transformer.
  arXiv  Detail & Related papers  (2020-06-10T08:22:41Z)

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.