mlsanityData to 14 Oct 2026

BAFPN: Bi directional alignment of features to improve localization accuracy

arXiv 2412.01859 ↗First posted 1 Dec 2024cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 29 of its 38 references that are in the map. First posted on arXiv 1 Dec 2024.

Builds on

29 of 38 references.

References · 29

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2016Feature Pyramid Networks for Object Detectioncited by 5,036
  3. 2017Squeeze-and-Excitation Networkscited by 4,789
  4. 2018CBAM: Convolutional Block Attention Modulecited by 2,477
  5. 2019FCOS: Fully Convolutional One-Stage Object Detectioncited by 1,577
  6. 2017Deformable Convolutional Networkscited by 1,488
  7. 2018Path Aggregation Network for Instance Segmentationcited by 798
  8. 2017DOTA: A Large-scale Dataset for Object Detection in Aerial Imagescited by 623
  9. 2019NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detectioncited by 291
  10. 2021Oriented R-CNN for Object Detectioncited by 231
  11. 2020DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolutioncited by 222
  12. 2017Arbitrary-Oriented Scene Text Detection via Rotation Proposalscited by 176
17 more
  1. 2021Rethinking Rotated Object Detection with Gaussian Wasserstein Distance Losscited by 156
  2. 2019Gliding vertex on the horizontal bounding box for multi-oriented object detectioncited by 134
  3. 2023Large Selective Kernel Network for Remote Sensing Object Detectioncited by 98
  4. 2021FaPN: Feature-aligned Pyramid Network for Dense Image Predictioncited by 81
  5. 2019Learning Modulated Loss for Rotated Object Detectioncited by 73
  6. 2019AugFPN: Improving Multi-scale Feature Learning for Object Detectioncited by 55
  7. 2023Adaptive Rotated Convolution for Rotated Object Detectioncited by 51
  8. 2020Effective Fusion Factor in FPN for Tiny Object Detectioncited by 44
  9. 2022H2RBox: Horizontal Box Annotation is All You Need for Oriented Object Detectioncited by 36
  10. 2024LR-FPN: Enhancing Remote Sensing Object Detection with Location Refined Feature Pyramid Networkcited by 27
  11. 2023AFPN: Asymptotic Feature Pyramid Network for Object Detectioncited by 17
  12. 2021A^2-FPN: Attention Aggregation based Feature Pyramid Network for Instance Segmentationcited by 15
  13. 2021CE-FPN: Enhancing Channel Information for Object Detectioncited by 8
  14. 2023Retro-FPN: Retrospective Feature Pyramid Network for Point Cloud Semantic Segmentationcited by 6
  15. 2020Cross-layer Feature Pyramid Network for Salient Object Detectioncited by 4
  16. 2021TricubeNet: 2D Kernel-Based Object Representation for Weakly-Occluded Oriented Object Detectioncited by 2
  17. 2022ssFPN: Scale Sequence (S^2) Feature Based-Feature Pyramid Network for Object Detectioncited by 2

Built on by

Not cited yet.