Digging Deep into the Layers of CNNs: In Search of How CNNs Achieve View Invariance

By Amr Bakry, Mohamed Elhoseiny, Tarek El-Gaaly, Ahmed M. Elgammal. 2015

31 citations recorded by the source corpus.

This paper is focused on studying the view-manifold structure in the feature spaces implied by the different layers of Convolutional Neural Networks (CNN). There are several questions that this paper aims to answer: Does the learned CNN representation achieve viewpoint invariance? How does it achieve viewpoint invariance? Is it achieved by collapsing the view manifolds, or separating them while preserving them? At which layer is view invariance achieved? How can the structure of the view manifold at each layer of a deep convolutional neural network be quantified experimentally? How does fine-tuning of a pre-trained CNN on a multi-view dataset affect the representation at each layer of the network? In order to answer these questions we propose a methodology to quantify the deformation and degeneracy of view manifolds in CNN layers. We apply this methodology and report interesting results in this paper that answer the aforementioned questions.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

What makes ImageNet good for transfer learning? (2016, background): “Another line of prior work investigated questions such as how distributed are CNN representations [2, 43], what is the the effect of feature transformations such as binarization [2, 14], how do CNN features change with systematic variations in scene factors [3], how invariant are features to differe”
Geometric Robustness of Deep Networks: Analysis and Improvement (2017, unknown): “The work in [2] uses the information about human visual system to understand and improve the transformation invariance.”
Overcoming Negative Transfer: A Survey (2020, background): “of domain discrepancy in traditional TL. HSIC is a nonparametric measure of the dependency between two sets, but it is not scale invariant and not designed to compare data in different feature spaces [85]. The Bregman divergence requires density estimation, which limits its applicability. MMD [6]–[8”
Convolutional Models for Joint Object Categorization and Pose Estimation (2015, methodology): “This in-depth study corroborates the conclusions we make here.”
iLab-20M: A Large-Scale Controlled Object Dataset to Investigate Deep Learning (2016, background): “[2] where they analytically found that fully connected layers make effort to collapse the low-dimensional intrinsic parameter manifolds to achieve invariant representations.”