UAV-DETR: DETR for Anti-Drone Target Detection

Jun Yang; Dong Wang; Hongxu Yin; Hongpeng Li; Jianxiong Yu

cs.CVMar 24, 2026

UAV-DETR: DETR for Anti-Drone Target Detection

Authors: Jun Yang, Dong Wang, Hongxu Yin, Hongpeng Li, Jianxiong Yu

Organizations: School of Automation, NPU, Xian, ShaanXi

Abstract

Drone detection is pivotal in numerous security and counter-UAV applications. However, existing deep learning-based methods typically struggle to balance robust feature representation with computational efficiency. This challenge is particularly acute when detecting miniature drones against complex backgrounds under severe environmental interference. To address these issues, we introduce UAV-DETR, a novel framework that integrates a small-target-friendly architecture with real-time detection capabilities. Specifically, UAV-DETR features a WTConv-enhanced backbone and a Sliding Window Self-Attention (SWSA-IFI) encoder, capturing the high-frequency structural details of tiny targets while drastically reducing parameter overhead. Furthermore, we propose an Efficient Cross-Scale Feature Recalibration and Fusion Network (ECFRFN) to suppress background noise and aggregate multi-scale semantics. To further enhance accuracy, UAV-DETR incorporates a hybrid Inner-CIoU and NWD loss strategy, mitigating the extreme sensitivity of standard IoU metrics to minor positional deviations in small objects. Extensive experiments demonstrate that UAV-DETR significantly outperforms the baseline RT-DETR on our custom UAV dataset (+6.61% in mAP50:95, with a 39.8% reduction in parameters) and the public DUT-ANTI-UAV benchmark (+1.4% in Precision, +1.0% in F1-Score). These results establish UAV-DETR as a superior trade-off between efficiency and precision in counter-UAV object detection. The code is available at https://github.com/wd-sir/UAVDETR.

Figures & tables

Figure 2: The process of the WTConv operation on a single channel, utilizing a 2-level wavelet decomposition and $5\times 5$ kernel sizes for the depth-wise convolutions.

Figure 3: (a) The detailed architecture of the proposed WTConv Block. (b) The structural principle of the SWSA attention mechanism.

Figure 4: (a) The architecture of the Selective Boundary Aggregation (SBA) module. (b) The schematic of the RepNCSPELAN4 module illustrating the structural re-parameterization mechanism.

Component	Specification
Operating System	Ubuntu 20.04.6 LTS
CPU	12th Gen Intel(R) Core(TM) i7-12700KF @ 3.60 GHz
RAM	64 GB
GPU	NVIDIA RTX3090 (24 GB)
Python Version	3.9.25
Deep Learning Framework	PyTorch 1.12.1

Table 1: Implementation Environment Details

Panel of Figure 6 — Figure 6: Qualitative comparison on the UAV dataset. The proposed UAV-DETR demonstrates superior stability and coverage in both (a) F1-Confidence and (b) Precision-Recall metrics.

Figure 7: Comparison of Model Performance vs. Parameters. The x-axis and y-axis denote F1-Score and mAP 50:95 , respectively, while the bubble size represents the number of parameters. Our UAV-DETR achieves the best trade-off, located in the top-right corner with a compact model size.

Model	$\boldsymbol{P}$ (%) $\uparrow$	$\boldsymbol{R}$ (%) $\uparrow$	$\boldsymbol{F1}$ (%) $\uparrow$	mAP 50 (%) $\uparrow$	mAP 75 (%) $\uparrow$	mAP 50:95 (%) $\uparrow$	FLOPs (G) $\downarrow$	Params $\downarrow$
Faster R-CNN ${}_{\text{PT}}$	53.65	47.51	50.40	43.62	6.04	14.41	401.7	136,689,024
SSD ${}_{\text{PT}}$	95.39	25.90	40.74	78.16	23.23	33.43	58.4	11,671,638
YOLOv8m	94.74	87.43	90.94	93.06	67.54	60.21	78.7	25,840,339
YOLOv10m	93.49	86.23	89.72	92.91	67.51	59.85	58.9	15,313,747
YOLO11m	94.51	87.50	90.87	92.97	67.02	59.69	67.6	20,030,803
YOLO12m	91.70	83.79	87.57	90.35	56.43	52.76	67.1	20,105,683

Table 2: Quantitative comparison of detection performance on the custom UAV dataset. The best results are highlighted in bold .

Figure 8: Visualization of detection results in an open field scenario. Purple and red boxes indicate missed detections and false alarms, respectively. The proposed UAV-DETR (bottom right) achieves precise localization of UAV targets with zero missed detections or false alarms..

Figure 9: Visualization of detection results against a cloudy mountainous sky. The proposed UAV-DETR (bottom right) exclusively achieves flawless localization with elevated confidence scores, successfully avoiding both the bird-induced false alarms common in most baselines and other environmental clutter.

Figure 10: Visualization of detection results in a complex environment with heavy tree occlusion and structural clutter. The severe visual fragmentation causes a complete detection failure across all tested YOLO variants. In contrast, the proposed UAV-DETR (bottom right) successfully penetrates the foliage clutter, achieving accurate localization with the highest confidence score of 0.80.

Figure 11: Sequential feature heatmap visualization demonstrating the internal data flow of UAV-DETR. From left to right: (1) raw inputs, (2) low-level features, (3) WTConv-enhanced backbone outputs, (4) SWSA-IFI encoder outputs, and (5) final ECFRFN neck activations. The visual progression highlights the robust suppression of background clutter and the precise accentuation of miniature targets.

Figure 12: Performance comparison on the DUT Anti-UAV dataset. UAV-DETR maintains its leading position, demonstrating strong generalization capabilities.

Model	$\boldsymbol{P}$ (%) $\uparrow$	$\boldsymbol{R}$ (%) $\uparrow$	$\boldsymbol{F1}$ (%) $\uparrow$	mAP 50 (%) $\uparrow$	mAP 75 (%) $\uparrow$	mAP 50:95 (%) $\uparrow$	FLOPs (G) $\downarrow$	Params $\downarrow$
Faster R-CNN ${}_{\text{PT}}$	77.62	54.83	64.27	49.89	25.82	26.01	401.7	136,689,024
SSD ${}_{\text{PT}}$	95.74	47.04	63.08	80.12	54.64	50.50	58.4	11,671,638
YOLOv8m	92.84	84.38	88.41	90.61	68.59	61.66	78.7	25,840,339
YOLOv10m	92.63	83.21	87.66	91.14	69.62	62.25	58.9	15,313,747
YOLO11m	94.21	84.14	88.89	90.99	70.37	62.01	67.6	20,030,803
YOLO12m	93.03	81.11	86.67	90.48	71.43	63.19	67.1	20,105,683

Table 3: Quantitative comparison of detection performance on the public DUT-ANTI-UAV benchmark. The best results are highlighted in bold .

Figure 13: Progressive ablation study evaluating the trade-off between detection accuracy and model complexity. The bar chart (left y-axis) denotes the mAP 50:95 metric ( $\uparrow$ ), while the line graph (right y-axis) tracks the model parameter count ( $\downarrow$ ). The incremental steps demonstrate the impact of each component, with the final configuration (Model O+A+B+C+D) representing our proposed UAV-DETR, achieving the optimal balance between high precision and lightweight design.

Model	$\boldsymbol{P}$ (%) $\uparrow$	$\boldsymbol{R}$ (%) $\uparrow$	$\boldsymbol{F1}$ (%) $\uparrow$	mAP 50 (%) $\uparrow$	mAP 75 (%) $\uparrow$	mAP 50:95 (%) $\uparrow$	FLOPs (G) $\downarrow$	Params $\downarrow$
O	96.28	93.63	94.94	95.45	58.55	55.95	56.9	19,873,044
O+A	95.95	94.18	95.05	95.76	62.71	58.01	56.9	19,873,044
O+A+B	96.51	94.60	95.54	96.06	67.61	60.61	79.7	18,038,324
O+A+B+C	96.22	94.44	95.32	95.54	65.03	59.19	65.2	12,082,484
O+A+B+D	96.81	94.89	95.84	96.23	68.74	61.40	79.7	17,842,620
O+A+B+C+D	96.82	94.93	95.87	96.58	71.08	62.56	66.7	11,962,040

Table 4: Ablation Study. The best results are highlighted in bold .

Panel of Figure 14 — Figure 14: Representative failure cases visualized by UAV-DETR on the custom UAV dataset. (a) FP where morphological similarities with birds lead to misclassification. (b) FN caused by miniature targets visually blending into complex urban architectural backgrounds.

Explore similar work

Feb 26, 2026

UFO-DETR: Frequency-Guided End-to-End Detector for UAV Tiny Objects

Small target detection in UAV imagery faces significant challenges such as scale variations, dense distribution, and the dominance of small targets. Existing algorithms rely on manually designed components, and general-purpose detectors are not optimized for UAV images, making it difficult to balance accuracy and complexity. To address these challenges, this paper proposes an end-to-end object detection framework, UFO-DETR, which integrates an LSKNet-based backbone network to optimize the receptive field and reduce the number of parameters. By combining the DAttention and AIFI modules, the model flexibly models multi-scale spatial relationships, improving multi-scale target detection performance. Additionally, the DynFreq-C3 module is proposed to enhance small target detection capability through cross-space frequency feature enhancement. Experimental results show that, compared to RT-DETR-L, the proposed method offers significant advantages in both detection performance and computational efficiency, providing an efficient solution for UAV edge computing.

Yuankai Chen, Kai Lin, Qihong Wu +6

Easy Lab, College of Mathematics and Informatics, South China Agricultural University, Guangzhou, China · School of Computer Science and Engineering, Southeast University, Nanjing, China · School of Mathematics, Foshan University, Foshan, China

Small Object Detection Aerial Object Detection Efficient Object Detection Detection Transformer Frequency-Domain Feature Learning

Ask PDF

Dec 19, 2025

YolovN-CBi: A Lightweight and Efficient Architecture for Real-Time Detection of Small UAVs

Unmanned Aerial Vehicles, commonly known as, drones pose increasing risks in civilian and defense settings, demanding accurate and real-time drone detection systems. However, detecting drones is challenging because of their small size, rapid movement, and low visual contrast. A modified architecture of YolovN called the YolovN-CBi is proposed that incorporates the Convolutional Block Attention Module (CBAM) and the Bidirectional Feature Pyramid Network (BiFPN) to improve sensitivity to small object detections. A curated training dataset consisting of 28K images is created with various flying objects and a local test dataset is collected with 2500 images consisting of very small drone objects. The proposed architecture is evaluated on four benchmark datasets, along with the local test dataset. The baseline Yolov5 and the proposed Yolov5-CBi architecture outperform newer Yolo versions, including Yolov8 and Yolov12, in the speed-accuracy trade-off for small object detection. Four other variants of the proposed CBi architecture are also proposed and evaluated, which vary in the placement and usage of CBAM and BiFPN. These variants are further distilled using knowledge distillation techniques for edge deployment, using a Yolov5m-CBi teacher and a Yolov5n-CBi student. The distilled model achieved a [email protected]:0.9 of 0.6573, representing a 6.51% improvement over the teacher's score of 0.6171, highlighting the effectiveness of the distillation process. The distilled model is 82.9% faster than the baseline model, making it more suitable for real-time drone detection. These findings highlight the effectiveness of the proposed CBi architecture, together with the distilled lightweight models in advancing efficient and accurate real-time detection of small UAVs.

Ami Pandat, Punna Rajasekhar, Gopika Vinod +1

Homi Bhabha National Institute, Mumbai, 400094, Maharashtra, India. · Bhabha Atomic Research Center, Mumbai, 400094, Maharashtra, India.

Small Object Detection Object Detection Real-Time Object Detection Aerial Object Detection

Ask PDF

Apr 14, 2026

DroneScan-YOLO: Redundancy-Aware Lightweight Detection for Tiny Objects in UAV Imagery

Aerial object detection in UAV imagery presents unique challenges due to the high prevalence of tiny objects, adverse environmental conditions, and strict computational constraints. Standard YOLO-based detectors fail to address these jointly: their minimum detection stride of 8 pixels renders sub-32px objects nearly undetectable, their CIoU loss produces zero gradients for non-overlapping tiny boxes, and their architectures contain significant filter redundancy. We propose DroneScan-YOLO, a holistic system contribution that addresses these limitations through four coordinated design choices: (1) increased input resolution of 1280x1280 to maximize spatial detail for tiny objects, (2) RPA-Block, a dynamic filter pruning mechanism based on lazy cosine-similarity updates with a 10-epoch warm-up period, (3) MSFD, a lightweight P2 detection branch at stride 4 adding only 114,592 parameters (+1.1%), and (4) SAL-NWD, a hybrid loss combining Normalized Wasserstein Distance with size-adaptive CIoU weighting, integrated into YOLOv8's TaskAligned assignment pipeline. Evaluated on VisDrone2019-DET, DroneScan-YOLO achieves 55.3% mAP@50 and 35.6% mAP@50-95, outperforming the YOLOv8s baseline by +16.6 and +12.3 points respectively, improving recall from 0.374 to 0.518, and maintaining 96.7 FPS inference speed with only +4.1% parameters. Gains are most pronounced on tiny object classes: bicycle AP@50 improves from 0.114 to 0.328 (+187%), and awning-tricycle from 0.156 to 0.237 (+52%).

Yann V. Bellec

Small Object Detection Aerial Object Detection Efficient Object Detection Structured Pruning

Ask PDF