[ICLR 2026] RF-DETR is a real-time object detection and segmentation model architecture developed by Roboflow, SOTA on COCO, designed for fine-tuning.

NVIDIA DeepStream SDK 8.0 / 7.1 / 7.0 / 6.4 / 6.3 / 6.2 / 6.1.1 / 6.1 / 6.0.1 / 6.0 application for YOLO-Segmentation models

C++ framework for computer vision inference, supporting multiple vision tasks and deep learning backends.

Object tracking pipelines complete with RF-DETR, YOLOv9, YOLO-NAS, YOLOv8, and YOLOv7 detection and BYTETracker tracking

ONNX model with inference, conversion and visualization scripts for RF-DETR (object detection and instance segmentation)

RF-DETR C++ tensorrt : Real-Time End-to-End Object Detection

C++ application to perform computer vision tasks using Nvidia Triton Server for model inference

RF-DETR Object Detection with DeepSORT Tracking

Run RF-DETR on NVIDIA DeepStream

RF-DETR + USLS: object detection using Rust

One‑line fine‑tuning of RF‑DETR on selected classes from OpenImages V7

RF-DETR wrapper for ROS 2 C++ implementation.

Advanced real-time football tracking system using RF-DETR, Extended Kalman Filter, and an intelligent virtual camera. Production-grade pipeline for sports broadcasting with low-latency RTMP/YouTube streaming, PID-controlled smoothing, and trajectory prediction.

RF-DETR C++ inference engine for object detection and instance segmentation with ONNX Runtime and TensorRT support

RF-DETR target model for use with Autodistill.

Fine-tune the SOTA RF-DETR model in AWS SageMaker or locally

FashionVeil: fashion apparel dataset with occlusion level annotations for clothes and accessories

Since have been noticed that Roboflow's RFDETRBase incorporates the drone class as an additional class, this drone detection program is presented.

RF-DETR for Docment Layout Analysis

Pose-based Taekwondo Poomsae Recognition with Geometric Normalization and Classical ML Baselines