FBRT-YOLO: Faster and Better for Real-Time Aerial Image Detection

Authors: Yao Xiao, Tingfa Xu, Yu Xin, Jianan Li

AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experimental results on three major aerial image datasets, including Visdrone, UAVDT, and AI-TOD, demonstrate that FBRT-YOLO outperforms various real-time detectors in terms of performance and speed. Experiments Implementation Details We conduct extensive experiments on three object detection benchmarks based on aerial images, i.e. Visdrone, UAVDT, and AI-TOD.
Researcher Affiliation Academia Yao Xiao, Tingfa Xu*, Yu Xin, Jianan Li Beijing Institute of Technology EMAIL
Pseudocode No The paper describes the proposed modules (FCM and MKP) with mathematical formulations and architectural diagrams (e.g., Fig. 3), but it does not include a clearly labeled pseudocode or algorithm block with structured steps.
Open Source Code Yes Code https://github.com/galaxy-oss/FCM.
Open Datasets Yes Extensive experimental results on three major aerial image datasets, including Visdrone, UAVDT, and AI-TOD, demonstrate that FBRT-YOLO outperforms various real-time detectors in terms of performance and speed. Extensive experiments conducted on widely-used aerial image benchmarks such as Vis Drone (2018), UAVDT (2018), and AI-TOD (2021) demonstrate that our FBRT-YOLO significantly outperforms previous state-of-the-art YOLO series models.
Dataset Splits No The paper mentions using well-known benchmark datasets such as Visdrone, UAVDT, and AI-TOD, but it does not explicitly specify the training, validation, and test splits (e.g., percentages or exact counts) used for these datasets in the provided text.
Hardware Specification Yes All experiments are conducted on an NVIDIA Ge Force RTX 4090 GPU, except that the inference speed is test on a single RTX 3080 GPU.
Software Dependencies No The paper mentions various model architectures like YOLOv8, YOLOv10, and RT-DETR, and training parameters like SGD optimizer, but it does not specify explicit version numbers for software dependencies such as Python, PyTorch, or CUDA.
Experiment Setup Yes Our network is trained for 300 epochs using the stochastic gradient descent (SGD) optimizer with a momentum of 0.937, a weight decay of 0.0005, a batch size of 4, and an initial learning rate of 0.01.