Deep roto-translation scattering for object classification

Oyallon, Edouard; Mallat, Stéphane

doi:10.1109/cvpr.2015.7298904

Cited by 157 publications

(188 citation statements)

References 23 publications

Supporting

Mentioning

187

Contrasting

Unclassified

Order By: Relevance

“…It is worth pointing out on CIFAR-10, the other reported methods [22,7,26,21,10] are usually based on different unsupervised networks and supervised classifiers for evaluation, making it difficult to make a direct comparison between them. The results still suggest that the state-of-the-art performances can be reached by AETs, as their error rates are very close to the pre-assumptive lower bound set by the fully supervised counterpart.…”

Section: Resultsmentioning

confidence: 99%

AET vs. AED: Unsupervised Representation Learning by Auto-Encoding Transformations Rather Than Data

Zhang¹,

Qi²,

Wang

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

206

180

View full text Add to dashboard Cite

The success of deep neural networks often relies on a large amount of labeled examples, which can be difficult to obtain in many real scenarios. To address this challenge, unsupervised methods are strongly preferred for training neural networks without using any labeled data. In this paper, we present a novel paradigm of unsupervised representation learning by Auto-Encoding Transformation (AET) in contrast to the conventional Auto-Encoding Data (AED) approach. Given a randomly sampled transformation, AET seeks to predict it merely from the encoded features as accurately as possible at the output end. The idea is the following: as long as the unsupervised features successfully encode the essential information about the visual structures of original and transformed images, the transformation can be well predicted. We will show that this AET paradigm allows us to instantiate a large variety of transformations, from parameterized, to non-parameterized and GAN-induced ones. Our experiments show that AET greatly improves over existing unsupervised approaches, setting new state-of-the-art performances being greatly closer to the upper bounds by their fully supervised counterparts on CIFAR-10, ImageNet and Places datasets.

show abstract

Section: Resultsmentioning

confidence: 99%

AET vs. AED: Unsupervised Representation Learning by Auto-Encoding Transformations Rather Than Data

Zhang¹,

Qi²,

Wang

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

206

180

View full text Add to dashboard Cite

show abstract

“…Roto-Scat + SVM [28] 17.7 ExamplarCNN [11] 15.7 DCGAN [32] 17.2 Scattering [27] 15.3 RotNet + non-linear [14] 10.94 RotNet + conv [14] 8.84 AET-affine + non-linear [37] 9.77 AET-affine + conv [37] 8.05 AET + non-linear [37] 9.41 AET + conv [37] 7.82 AVT + non-linear 8.96 AVT + conv 7.75 Table 2: Error rates of different classifiers trained on top of the learned representations on CIFAR 10, where n-FC denotes a classifier with n fully connected layers and conv denotes the third NIN block as a convolutional classifier. Two AET variants are chosen for a fair direct comparison since they are based on the same architecture as the AVT and have outperformed the other unsupervised representations before [37].…”

Section: Methodsmentioning

confidence: 99%

AVT: Unsupervised Learning of Transformation Equivariant Representations by Autoencoding Variational Transformations

Zhang

Chen

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

View full text Add to dashboard Cite

The learning of Transformation-Equivariant Representations (TERs), which is introduced by Hinton et al. [16], has been considered as a principle to reveal visual structures under various transformations. It contains the celebrated Convolutional Neural Networks (CNNs) as a special case that only equivary to the translations. In contrast, we seek to train TERs for a generic class of transformations and train them in an unsupervised fashion. To this end, we present a novel principled method by Autoencoding Variational Transformations (AVT), compared with the conventional approach to autoencoding data. Formally, given transformed images, the AVT seeks to train the networks by maximizing the mutual information between the transformations and representations. This ensures the resultant TERs of individual images contain the intrinsic information about their visual structures that would equivary extricably under various transformations in a generalized nonlinear case. Technically, we show that the resultant optimization problem can be efficiently solved by maximizing a variational lower-bound of the mutual information. This variational approach introduces a transformation decoder to approximate the intractable posterior of transformations, resulting in an autoencoding architecture with a pair of the representation encoder and the transformation decoder. Experiments demonstrate the proposed AVT model sets a new record for the performances on unsupervised tasks, greatly closing the performance gap to the supervised models.

show abstract

“…Recently there has been an explosion of interest into CNNs with predefined transformation equivariances, beyond translation [2,3,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19] [20, 21,22,23]. However, with the exception of Cohen and Welling [23] (projections on sphere), Kondor [22] (point clouds), and Thomas et al [20] (point clouds), these have mainly focused on the 2D scenario.…”

Section: Related Workmentioning

confidence: 99%

CubeNet: Equivariance to 3D Rotation and Translation

Worrall

Brostow

2018

Computer Vision – ECCV 2018

109

106

View full text Add to dashboard Cite

3D Convolutional Neural Networks are sensitive to transformations applied to their input. This is a problem because a voxelized version of a 3D object, and its rotated clone, will look unrelated to each other after passing through to the last layer of a network. Instead, an idealized model would preserve a meaningful representation of the voxelized object, while explaining the pose-difference between the two inputs. An equivariant representation vector has two components: the invariant identity part, and a discernable encoding of the transformation. Models that can't explain pose-differences risk "diluting" the representation, in pursuit of optimizing a classification or regression loss function.We introduce a Group Convolutional Neural Network with linear equivariance to translations and right angle rotations in three dimensions. We call this network CubeNet, reflecting its cube-like symmetry. By construction, this network helps preserve a 3D shape's global and local signature, as it is transformed through successive layers. We apply this network to a variety of 3D inference problems, achieving state-of-the-art on the ModelNet10 classification challenge, and comparable performance on the ISBI 2012 Connectome Segmentation Benchmark. To the best of our knowledge, this is the first 3D rotation equivariant CNN for voxel representations.

show abstract

Deep roto-translation scattering for object classification

Cited by 157 publications

References 23 publications

AET vs. AED: Unsupervised Representation Learning by Auto-Encoding Transformations Rather Than Data

AET vs. AED: Unsupervised Representation Learning by Auto-Encoding Transformations Rather Than Data

AVT: Unsupervised Learning of Transformation Equivariant Representations by Autoencoding Variational Transformations

CubeNet: Equivariance to 3D Rotation and Translation

Contact Info

Product

Resources

About