Content-Adaptive Downsampling in Convolutional Neural Networks

• URL: http://arxiv.org/abs/2305.09504v1
• Date: Tue, 16 May 2023 14:58:30 GMT
• Title: Content-Adaptive Downsampling in Convolutional Neural Networks
• Authors: Robin Hesse, Simone Schaub-Meyer, Stefan Roth
• Abstract summary: Many convolutional neural networks (CNNs) rely on progressive downsampling of their feature maps to increase the network's receptive field and decrease computational cost. This comes at the price of losing granularity in the feature maps, limiting the ability to correctly understand images or recover fine detail in dense prediction tasks. We propose an adaptive downsampling scheme that generalizes the above idea by allowing to process informative regions at a higher resolution than less informative ones.
• Score: 15.073405675079558
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Many convolutional neural networks (CNNs) rely on progressive downsampling of their feature maps to increase the network's receptive field and decrease computational cost. However, this comes at the price of losing granularity in the feature maps, limiting the ability to correctly understand images or recover fine detail in dense prediction tasks. To address this, common practice is to replace the last few downsampling operations in a CNN with dilated convolutions, allowing to retain the feature map resolution without reducing the receptive field, albeit increasing the computational cost. This allows to trade off predictive performance against cost, depending on the output feature resolution. By either regularly downsampling or not downsampling the entire feature map, existing work implicitly treats all regions of the input image and subsequent feature maps as equally important, which generally does not hold. We propose an adaptive downsampling scheme that generalizes the above idea by allowing to process informative regions at a higher resolution than less informative ones. In a variety of experiments, we demonstrate the versatility of our adaptive downsampling strategy and empirically show that it improves the cost-accuracy trade-off of various established CNNs.

Related papers

• CEC-CNN: A Consecutive Expansion-Contraction Convolutional Network for Very Small Resolution Medical Image Classification [0.8108972030676009]
  We introduce a new CNN architecture which preserves multi-scale features from deep, intermediate, and shallow layers. Using a dataset of very low resolution patches from Pancreatic Ductal Adenocarcinoma (PDAC) CT scans we demonstrate that our network can outperform current state of the art models.
  arXiv  Detail & Related papers  (2022-09-27T20:01:12Z)
• Adaptive Self-supervision Algorithms for Physics-informed Neural Networks [59.822151945132525]
  Physics-informed neural networks (PINNs) incorporate physical knowledge from the problem domain as a soft constraint on the loss function. We study the impact of the location of the collocation points on the trainability of these models. We propose a novel adaptive collocation scheme which progressively allocates more collocation points to areas where the model is making higher errors.
  arXiv  Detail & Related papers  (2022-07-08T18:17:06Z)
• SAR Despeckling Using Overcomplete Convolutional Networks [53.99620005035804]
  despeckling is an important problem in remote sensing as speckle degrades SAR images. Recent studies show that convolutional neural networks(CNNs) outperform classical despeckling methods. This study employs an overcomplete CNN architecture to focus on learning low-level features by restricting the receptive field. We show that the proposed network improves despeckling performance compared to recent despeckling methods on synthetic and real SAR images.
  arXiv  Detail & Related papers  (2022-05-31T15:55:37Z)
• Unpaired Image Super-Resolution with Optimal Transport Maps [128.1189695209663]
  Real-world image super-resolution (SR) tasks often do not have paired datasets limiting the application of supervised techniques. We propose an algorithm for unpaired SR which learns an unbiased OT map for the perceptual transport cost. Our algorithm provides nearly state-of-the-art performance on the large-scale unpaired AIM-19 dataset.
  arXiv  Detail & Related papers  (2022-02-02T16:21:20Z)
• Revisiting Point Cloud Simplification: A Learnable Feature Preserving Approach [57.67932970472768]
  Mesh and Point Cloud simplification methods aim to reduce the complexity of 3D models while retaining visual quality and relevant salient features. We propose a fast point cloud simplification method by learning to sample salient points. The proposed method relies on a graph neural network architecture trained to select an arbitrary, user-defined, number of points from the input space and to re-arrange their positions so as to minimize the visual perception error.
  arXiv  Detail & Related papers  (2021-09-30T10:23:55Z)
• Dynamic Resolution Network [40.64164953983429]
  The redundancy on the input resolution of modern CNNs has not been fully investigated. We propose a novel dynamic-resolution network (DRNet) in which the resolution is determined dynamically based on each input sample. DRNet achieves similar performance with an about 34% reduction, while gains 1.4% accuracy increase with 10% reduction compared to the original ResNet-50 on ImageNet.
  arXiv  Detail & Related papers  (2021-06-05T13:48:33Z)
• Truly shift-invariant convolutional neural networks [0.0]
  Recent works have shown that the output of a CNN can change significantly with small shifts in input. We propose adaptive polyphase sampling (APS), a simple sub-sampling scheme that allows convolutional neural networks to achieve 100% consistency in classification performance under shifts.
  arXiv  Detail & Related papers  (2020-11-28T20:57:35Z)
• Beyond Dropout: Feature Map Distortion to Regularize Deep Neural Networks [107.77595511218429]
  In this paper, we investigate the empirical Rademacher complexity related to intermediate layers of deep neural networks. We propose a feature distortion method (Disout) for addressing the aforementioned problem. The superiority of the proposed feature map distortion for producing deep neural network with higher testing performance is analyzed and demonstrated.
  arXiv  Detail & Related papers  (2020-02-23T13:59:13Z)
• Image Fine-grained Inpainting [89.17316318927621]
  We present a one-stage model that utilizes dense combinations of dilated convolutions to obtain larger and more effective receptive fields. To better train this efficient generator, except for frequently-used VGG feature matching loss, we design a novel self-guided regression loss. We also employ a discriminator with local and global branches to ensure local-global contents consistency.
  arXiv  Detail & Related papers  (2020-02-07T03:45:25Z)
• Improving the Resolution of CNN Feature Maps Efficiently with Multisampling [8.655380837944188]
  One version of our method, which we call subsampling, significantly improves the accuracy of state-of-the-art architectures such as DenseNet and ResNet without any additional parameters. We glean possible insight into the nature of data augmentations and demonstrate experimentally that coarse feature maps are bottlenecking the performance of neural networks in image classification.
  arXiv  Detail & Related papers  (2018-05-28T04:29:02Z)

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.