BAFPN: Bi directional alignment of features to improve localization accuracy

TL;DR

The paper introduces BAFPN, a novel feature pyramid network that globally aligns features to enhance high-precision object localization, outperforming traditional methods on the DOTAv1.5 dataset.

Contribution

It proposes a bidirectional alignment approach with SPAM and SEAM modules to address spatial misalignment and aliasing in feature pyramids, improving localization accuracy.

Findings

Improves AP75 by 1.68% on DOTAv1.5

Enhances AP50 by 1.45%

Boosts mAP by 1.34%

Abstract

Current state-of-the-art vision models often utilize feature pyramids to extract multi-scale information, with the Feature Pyramid Network (FPN) being one of the most widely used classic architectures. However, traditional FPNs and their variants (e.g., AUGFPN, PAFPN) fail to fully address spatial misalignment on a global scale, leading to suboptimal performance in high-precision localization of objects. In this paper, we propose a novel Bidirectional Alignment Feature Pyramid Network (BAFPN), which aligns misaligned features globally through a Spatial Feature Alignment Module (SPAM) during the bottom-up information propagation phase. Subsequently, it further mitigates aliasing effects caused by cross-scale feature fusion via a fine-grained Semantic Alignment Module (SEAM) in the top-down phase. On the DOTAv1.5 dataset, BAFPN improves the baseline model's AP75, AP50, and mAP by 1.68%, 1.45%, and 1.34%, respectively. Additionally, BAFPN demonstrates significant performance gains when applied to various other advanced detectors.