Cascade Network with Guided Loss and Hybrid Attention for Two-view Geometry

TL;DR

This paper introduces a cascade network for two-view geometry that combines a guided loss function aligned with Fn-measure optimization and a hybrid attention mechanism to enhance feature extraction, achieving state-of-the-art results.

Contribution

The paper proposes a novel guided loss function that directly correlates with Fn-measure and a hybrid attention block integrating BACN and CA for improved feature extraction.

Findings

Achieves state-of-the-art performance on benchmark datasets.

Demonstrates effective optimization of Fn-measure during training.

Shows improved feature extraction through hybrid attention mechanisms.

Abstract

In this paper, we are committed to designing a high-performance network for two-view geometry. We first propose a Guided Loss and theoretically establish the direct negative correlation between the loss and Fn-measure by dynamically adjusting the weights of positive and negative classes during training, so that the network is always trained towards the direction of increasing Fn-measure. By this way, the network can maintain the advantage of the cross-entropy loss while maximizing the Fn-measure. We then propose a hybrid attention block to extract feature, which integrates the bayesian attentive context normalization (BACN) and channel-wise attention (CA). BACN can mine the prior information to better exploit global context and CA can capture complex channel context to enhance the channel awareness of the network. Finally, based on our Guided Loss and hybrid attention block, a cascade network is designed to gradually optimize the result for more superior performance. Experiments have shown that our network achieves the state-of-the-art performance on benchmark datasets.