visdrone

icip2022 paper: sahi benchmark on visdrone and xview datasets using fcos, vfnet and tood detectors

ECCV2018(Challenge-Object Detection in Images)

VisDrone aerial object detection toolkit with 33 models (Torchvision + YOLO), training, evaluation, video inference, benchmarking, and annotation conversion.

Many yolov8 model are trained on the VisDrone dataset.

YOLO-TLP: detected and classified tiny objects with bounding box dimensions smaller than 15 pixels, outperforming other one-stage detectors. maximum resolution for target observation in real-time applications.

Simple implement of CenterNet on VisDrone dataset.

Hybrid RT DETR: Hybrid encoder-decoder network for end-to-end object detection in UAV imagery

Small Object Detection in Dense UAV Imagery using YOLOv8-L with Structured Ablation (EMA, P2 Head, PIoU) and SAHI Evaluation on VisDrone2019.

dataset or annotation file format conversion

An end-to-end computer vision system for aerial imagery, implemented and validated on a commercial drone for real-time video inference

Object Detection on the Visdrone dataset

UAV aerial object detection and tracking (YOLOv8n + SORT) on VisDrone Dataset, with ONNX FP32/INT8 export and CPU deployment benchmarking.

Small object detection on VisDrone with YOLOv11s + SAHI

Smart traffic analysis application.

Object detection format converter from VisDrone2019-DET to Yolo.

Vehicle Detection using YOLOv26 trained on VisDrone dataset for detecting cars,vans,trucks and buses

Reproducible VisDrone benchmark scaffold for CNN, DETR, Vision Mamba, and RT-DETR families.

Cross-architecture knowledge distillation from RT-DETR to YOLO26n for label-scarce aerial detection (VisDrone), with semi-supervised pseudo-labelling and accuracy-gated INT8 PTQ and TensorRT export.

An end-to-end edge computer vision pipeline running real-time YOLOv8 aerial object detection on aerial drone footage optimized on CoreML & Neural Engine. Includes multi-runtime benchmarking, CLI inference, and automated testing.