Training Image Derivatives: Increased Accuracy and Universal Robustness
- URL: http://arxiv.org/abs/2310.14045v2
- Date: Mon, 27 Nov 2023 19:43:36 GMT
- Title: Training Image Derivatives: Increased Accuracy and Universal Robustness
- Authors: Vsevolod I. Avrutskiy
- Abstract summary: Derivative training is a known method that significantly improves the accuracy of neural networks in some low-dimensional applications.
In this paper, a similar improvement is obtained for an image analysis problem: reconstructing the vertices of a cube from its image.
The derivatives also offer insight into the robustness problem, which is currently understood in terms of two types of network vulnerabilities.
- Score: 3.9160947065896803
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Derivative training is a known method that significantly improves the
accuracy of neural networks in some low-dimensional applications. In this
paper, a similar improvement is obtained for an image analysis problem:
reconstructing the vertices of a cube from its image. By training the
derivatives with respect to the 6 degrees of freedom of the cube, we obtain 25
times more accurate results for noiseless inputs. The derivatives also offer
insight into the robustness problem, which is currently understood in terms of
two types of network vulnerabilities. The first type involves small
perturbations that dramatically change the output, and the second type relates
to substantial image changes that the network erroneously ignores. Defense
against each is possible, but safeguarding against both while maintaining the
accuracy defies conventional training methods. The first type is analyzed using
the network's gradient, while the second relies on human input evaluation,
serving as an oracle substitute. For the task at hand, the nearest neighbor
oracle can be defined and expanded into Taylor series using image derivatives.
This allows for a robustness analysis that unifies both types of
vulnerabilities and enables training where accuracy and universal robustness
are limited only by network capacity.
Related papers
- WiNet: Wavelet-based Incremental Learning for Efficient Medical Image Registration [68.25711405944239]
Deep image registration has demonstrated exceptional accuracy and fast inference.
Recent advances have adopted either multiple cascades or pyramid architectures to estimate dense deformation fields in a coarse-to-fine manner.
We introduce a model-driven WiNet that incrementally estimates scale-wise wavelet coefficients for the displacement/velocity field across various scales.
arXiv Detail & Related papers (2024-07-18T11:51:01Z) - Uncertainty Propagation through Trained Deep Neural Networks Using
Factor Graphs [4.704825771757308]
Uncertainty propagation seeks to estimate aleatoric uncertainty by propagating input uncertainties to network predictions.
Motivated by the complex information flows within deep neural networks, we developed a novel approach by posing uncertainty propagation as a non-linear optimization problem using factor graphs.
arXiv Detail & Related papers (2023-12-10T17:26:27Z) - Enhancing Multiple Reliability Measures via Nuisance-extended
Information Bottleneck [77.37409441129995]
In practical scenarios where training data is limited, many predictive signals in the data can be rather from some biases in data acquisition.
We consider an adversarial threat model under a mutual information constraint to cover a wider class of perturbations in training.
We propose an autoencoder-based training to implement the objective, as well as practical encoder designs to facilitate the proposed hybrid discriminative-generative training.
arXiv Detail & Related papers (2023-03-24T16:03:21Z) - THAT: Two Head Adversarial Training for Improving Robustness at Scale [126.06873298511425]
We propose Two Head Adversarial Training (THAT), a two-stream adversarial learning network that is designed to handle the large-scale many-class ImageNet dataset.
The proposed method trains a network with two heads and two loss functions; one to minimize feature-space domain shift between natural and adversarial images, and one to promote high classification accuracy.
arXiv Detail & Related papers (2021-03-25T05:32:38Z) - Convex Regularization Behind Neural Reconstruction [21.369208659395042]
This paper advocates a convex duality framework to make neural networks amenable to convex solvers.
Experiments with MNIST fastMRI datasets confirm the efficacy of the dual network optimization problem.
arXiv Detail & Related papers (2020-12-09T16:57:16Z) - Encoding Robustness to Image Style via Adversarial Feature Perturbations [72.81911076841408]
We adapt adversarial training by directly perturbing feature statistics, rather than image pixels, to produce robust models.
Our proposed method, Adversarial Batch Normalization (AdvBN), is a single network layer that generates worst-case feature perturbations during training.
arXiv Detail & Related papers (2020-09-18T17:52:34Z) - Second Order Optimization for Adversarial Robustness and
Interpretability [6.700873164609009]
We propose a novel regularizer which incorporates first and second order information via a quadratic approximation to the adversarial loss.
It is shown that using only a single iteration in our regularizer achieves stronger robustness than prior gradient and curvature regularization schemes.
It retains the interesting facet of AT that networks learn features which are well-aligned with human perception.
arXiv Detail & Related papers (2020-09-10T15:05:14Z) - Adversarial Training Reduces Information and Improves Transferability [81.59364510580738]
Recent results show that features of adversarially trained networks for classification, in addition to being robust, enable desirable properties such as invertibility.
We show that the Adversarial Training can improve linear transferability to new tasks, from which arises a new trade-off between transferability of representations and accuracy on the source task.
arXiv Detail & Related papers (2020-07-22T08:30:16Z) - Uncertainty Quantification in Deep Residual Neural Networks [0.0]
Uncertainty quantification is an important and challenging problem in deep learning.
Previous methods rely on dropout layers which are not present in modern deep architectures or batch normalization which is sensitive to batch sizes.
We show that training residual networks using depth can be interpreted as a variational approximation to the posterior weights in neural networks.
arXiv Detail & Related papers (2020-07-09T16:05:37Z) - Hidden Cost of Randomized Smoothing [72.93630656906599]
In this paper, we point out the side effects of current randomized smoothing.
Specifically, we articulate and prove two major points: 1) the decision boundaries of smoothed classifiers will shrink, resulting in disparity in class-wise accuracy; 2) applying noise augmentation in the training process does not necessarily resolve the shrinking issue due to the inconsistent learning objectives.
arXiv Detail & Related papers (2020-03-02T23:37:42Z)
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.