# BFA-YOLO: A balanced multiscale object detection network for building faccade attachments detection

arXiv: https://arxiv.org/abs/2409.04025 · HTML: https://mlsanity.com/paper/2409.04025/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 57 references that are in the map.
- First posted on arXiv 6 Sep 2024.

## Builds on (most cited first, 9 of 9)

- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [Group Normalization](https://mlsanity.com/paper/1803.08494/) (2018)
- [Deformable ConvNets v2: More Deformable, Better Results](https://mlsanity.com/paper/1811.11168/) (2018)
- [Understanding the Effective Receptive Field in Deep Convolutional Neural Networks](https://mlsanity.com/paper/1701.04128/) (2017)
- [TOOD: Task-aligned One-stage Object Detection](https://mlsanity.com/paper/2108.07755/) (2021)
- [TIDE: A General Toolbox for Identifying Object Detection Errors](https://mlsanity.com/paper/2008.08115/) (2020)
- [RMT: Retentive Networks Meet Vision Transformers](https://mlsanity.com/paper/2309.11523/) (2023)
- [Analysis of Generalizability of Deep Neural Networks Based on the Complexity of Decision Boundary](https://mlsanity.com/paper/2009.07974/) (2020)
- [Improving Facade Parsing with Vision Transformers and Line Integration](https://mlsanity.com/paper/2309.15523/) (2023)

## Abstract (arXiv)

The detection of faccade elements on buildings, such as doors, windows, balconies, air conditioning units, billboards, and glass curtain walls, is a critical step in automating the creation of Building Information Modeling (BIM). Yet, this field faces significant challenges, including the uneven distribution of faccade elements, the presence of small objects, and substantial background noise, which hamper detection accuracy. To address these issues, we develop the BFA-YOLO model and the BFA-3D dataset in this study. The BFA-YOLO model is an advanced architecture designed specifically for analyzing multi-view images of faccade attachments. It integrates three novel components: the Feature Balanced Spindle Module (FBSM) that tackles the issue of uneven object distribution; the Target Dynamic Alignment Task Detection Head (TDATH) that enhances the detection of small objects; and the Position Memory Enhanced Self-Attention Mechanism (PMESA), aimed at reducing the impact of background noise. These elements collectively enable BFA-YOLO to effectively address each challenge, thereby improving model robustness and detection precision. The BFA-3D dataset, offers multi-view images with precise annotations across a wide range of faccade attachment categories. This dataset is developed to address the limitations present in existing faccade detection datasets, which often feature a single perspective and insufficient category coverage. Through comparative analysis, BFA-YOLO demonstrated improvements of 1.8% and 2.9% in mAP50 on the BFA-3D dataset and the public Faccade-WHU dataset, respectively, when compared to the baseline YOLOv8 model. These results highlight the superior performance of BFA-YOLO in faccade element detection and the advancement of intelligent BIM technologies.