TablePage.md

The table represents a specific model and includes the following details:

• Model: The name or identifier of the model.
• Model Type: The type of model, indicating the underlying architecture or framework used (also the string corresponding to the param to be used with --type during the inference step of this project).
• o: Supported backend.
• x: Not supported backend (or at least not tested yet/not tried to export).
• Note: OpenCV-DNN --> tested only onnx weights on cpu, opencv-dnn currently supports only static shapes, dynamic input (or layer) shape does not work with opencv

Model	Model Type	OpenCV-DNN	TensorRT	LibTorch	Onnx-runtime	OpenVino	Libtensorflow
yolov4/yolov4-tiny	yolov4	o	x	x	x	x	x
yolov5 models	yolo	o	x	o	o	o	x
yolov6 models	yolo	o	x	x	o	x	x
yolov7 models	yolo	o	x	o	o	x	x
yolov8 models	yolo	o	o	o	o	o	o
yolov9 models	yolo	o	o	o	o	x	x
yolov10 models	yolov10	x	o	o	o	o	o
yolo11 models	yolo	o	o	o	o	x	x
yolov12 models	yolo	o	o	o	o	o	o
yolo26 models	yolo26	x	o	o	o	o	o
yolo-nas models	yolonas	o	x	x	o	x	x
rt-detr v1/v2/v4 models	rtdetr	x	o	x	o	x	x
rt-detr ultralytics models	rtdetrul	x	o	o	o	x	x
d-fine models	rtdetr	x	o	x	o	x	x
deim v1/v2 models	rtdetr	x	o	x	o	x	x
rf-detr models	rfdetr	x	o	o	o	x	x

Classification

Model	Model Type	OpenCV-DNN	TensorRT	LibTorch	Onnx-runtime	OpenVino	Libtensorflow
TorchVision Models (ResNet, EfficientNet...)	torchvisionclassifier	o	o	o	o	o	x
TensorFlow Models (MobileNet, Inception...)	tensorflowclassifier	x	x	x	x	x	o
Vision Transformers (ViT, Swin...)	vitclassifier	x	o	o	o	x	x
Video Transformers (TimeSformer)	timesformer	x	x	o	o	x	x

Instance Segmentation

Model	Model Type	OpenCV-DNN	TensorRT	LibTorch	Onnx-runtime	OpenVino	Libtensorflow
YOLO Segmentation (v5, v8, v11)	yoloseg	x	o	o	o	o	x
RF-DETR Segmentation	rfdetr-seg	x	o	o	o	x	x

Optical Flow

Model	Model Type	OpenCV-DNN	TensorRT	LibTorch	Onnx-runtime	OpenVino	Libtensorflow
RAFT	raft	x	o	o	o	x	x