Enhanced Performance of Pre-Trained Networks by Matched Augmentation Distributions

• URL: http://arxiv.org/abs/2201.07894v1
• Date: Wed, 19 Jan 2022 22:33:00 GMT
• Title: Enhanced Performance of Pre-Trained Networks by Matched Augmentation Distributions
• Authors: Touqeer Ahmad, Mohsen Jafarzadeh, Akshay Raj Dhamija, Ryan Rabinowitz, Steve Cruz, Chunchun Li, Terrance E. Boult
• Abstract summary: We propose a simple solution to address the train-test distributional shift. We combine results for multiple random crops for a test image. This not only matches the train time augmentation but also provides the full coverage of the input image.
• Score: 10.74023489125222
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: There exists a distribution discrepancy between training and testing, in the way images are fed to modern CNNs. Recent work tried to bridge this gap either by fine-tuning or re-training the network at different resolutions. However re-training a network is rarely cheap and not always viable. To this end, we propose a simple solution to address the train-test distributional shift and enhance the performance of pre-trained models -- which commonly ship as a package with deep learning platforms \eg, PyTorch. Specifically, we demonstrate that running inference on the center crop of an image is not always the best as important discriminatory information may be cropped-off. Instead we propose to combine results for multiple random crops for a test image. This not only matches the train time augmentation but also provides the full coverage of the input image. We explore combining representation of random crops through averaging at different levels \ie, deep feature level, logit level, and softmax level. We demonstrate that, for various families of modern deep networks, such averaging results in better validation accuracy compared to using a single central crop per image. The softmax averaging results in the best performance for various pre-trained networks without requiring any re-training or fine-tuning whatsoever. On modern GPUs with batch processing, the paper's approach to inference of pre-trained networks, is essentially free as all images in a batch can all be processed at once.

Related papers

• Boosting Verified Training for Robust Image Classifications via Abstraction [20.656457368486876]
  This paper proposes a novel, abstraction-based, certified training method for robust image classifiers. By training on intervals, all perturbed images that are mapped to the same interval are classified as the same label. For the abstraction, our training method also enables a sound and complete black-box verification approach.
  arXiv  Detail & Related papers  (2023-03-21T02:38:14Z)
• Training Your Sparse Neural Network Better with Any Mask [106.134361318518]
  Pruning large neural networks to create high-quality, independently trainable sparse masks is desirable. In this paper we demonstrate an alternative opportunity: one can customize the sparse training techniques to deviate from the default dense network training protocols. Our new sparse training recipe is generally applicable to improving training from scratch with various sparse masks.
  arXiv  Detail & Related papers  (2022-06-26T00:37:33Z)
• CropMix: Sampling a Rich Input Distribution via Multi-Scale Cropping [97.05377757299672]
  We present a simple method, CropMix, for producing a rich input distribution from the original dataset distribution. CropMix can be seamlessly applied to virtually any training recipe and neural network architecture performing classification tasks. We show that CropMix is of benefit to both contrastive learning and masked image modeling towards more powerful representations.
  arXiv  Detail & Related papers  (2022-05-31T16:57:28Z)
• On Efficient Transformer and Image Pre-training for Low-level Vision [74.22436001426517]
  Pre-training has marked numerous state of the arts in high-level computer vision. We present an in-depth study of image pre-training. We find pre-training plays strikingly different roles in low-level tasks.
  arXiv  Detail & Related papers  (2021-12-19T15:50:48Z)
• Tensor Normalization and Full Distribution Training [3.962145079528281]
  pixel wise normalization, which is inserted after linear units and batch normalization, provides a significant improvement in the accuracy of modern deep neural networks. We show that the factorized superposition of images from the training set and the reformulation of the multi class problem into a multi-label problem yields significantly more robust networks.
  arXiv  Detail & Related papers  (2021-09-06T10:33:17Z)
• ResMLP: Feedforward networks for image classification with data-efficient training [73.26364887378597]
  We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. We will share our code based on the Timm library and pre-trained models.
  arXiv  Detail & Related papers  (2021-05-07T17:31:44Z)
• Jigsaw Clustering for Unsupervised Visual Representation Learning [68.09280490213399]
  We propose a new jigsaw clustering pretext task in this paper. Our method makes use of information from both intra- and inter-images. It is even comparable to the contrastive learning methods when only half of training batches are used.
  arXiv  Detail & Related papers  (2021-04-01T08:09:26Z)
• Twice Mixing: A Rank Learning based Quality Assessment Approach for Underwater Image Enhancement [42.03072878219206]
  We propose a rank learning guided no-reference quality assessment method for underwater image enhancement (UIE) Our approach, termed Twice Mixing, is motivated by the observation that a mid-quality image can be generated by mixing a high-quality image with its low-quality version. We conduct extensive experiments on both synthetic and real-world datasets.
  arXiv  Detail & Related papers  (2021-02-01T07:13:39Z)
• An Empirical Study of the Collapsing Problem in Semi-Supervised 2D Human Pose Estimation [80.02124918255059]
  Semi-supervised learning aims to boost the accuracy of a model by exploring unlabeled images. We learn two networks to mutually teach each other. The more reliable predictions on easy images in each network are used to teach the other network to learn about the corresponding hard images.
  arXiv  Detail & Related papers  (2020-11-25T03:29:52Z)
• Increasing the Robustness of Semantic Segmentation Models with Painting-by-Numbers [39.95214171175713]
  We build upon an insight from image classification that output can be improved by increasing the network-bias towards object shapes. Our basic idea is to alpha-blend a portion of the RGB training images with faked images, where each class-label is given a fixed, randomly chosen color. We demonstrate the effectiveness of our training schema for DeepLabv3+ with various network backbones, MobileNet-V2, ResNets, and Xception, and evaluate it on the Cityscapes dataset.
  arXiv  Detail & Related papers  (2020-10-12T07:42:39Z)
• DiverseNet: When One Right Answer is not Enough [35.764028730120096]
  We introduce a simple method for training a neural network, which enables diverse structured predictions to be made for each test-time query. Our method results in quantitative improvements across three challenging tasks: 2D image completion, 3D volume estimation, and flow prediction.
  arXiv  Detail & Related papers  (2020-08-24T18:12:49Z)

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.