Local Feature Descriptor Learning with Adaptive Siamese Network

TL;DR

This paper introduces an adaptive pruning Siamese network for local feature descriptor learning, improving efficiency and recognition accuracy by dynamically adjusting network size based on neuron activation.

Contribution

It proposes a novel adaptive pruning approach for Siamese networks that enhances local feature descriptor learning efficiency and accuracy.

Findings

Outperforms state-of-the-art in patch matching

More computationally efficient due to adaptive pruning

Improves recognition rate over complex networks

Abstract

Although the recent progress in the deep neural network has led to the development of learnable local feature descriptors, there is no explicit answer for estimation of the necessary size of a neural network. Specifically, the local feature is represented in a low dimensional space, so the neural network should have more compact structure. The small networks required for local feature descriptor learning may be sensitive to initial conditions and learning parameters and more likely to become trapped in local minima. In order to address the above problem, we introduce an adaptive pruning Siamese Architecture based on neuron activation to learn local feature descriptors, making the network more computationally efficient with an improved recognition rate over more complex networks. Our experiments demonstrate that our learned local feature descriptors outperform the state-of-art methods in patch matching.