[Model] Yolov5/v5lite/v6/v7/v7end2end: CUDA preprocessing

[wang-xinyu]

PR types

Performance optimization

PR changes

Others - preprocessing

Describe

• Add a YOLO CUDA preprocessing util
• Yolov5/v5lite/v6/v7/v7end2end: integrate CUDA preprocessing, test and compare latency
• cmake changes to support CUDA source files compile

[CLAassistant]

All committers have signed the CLA.

[wang-xinyu]

Latency includes preprocessing, inference and postprocessing, in milliseconds.
Tested on P40, TensorRT8.4.

Model	Latency(CPU preprocessing)	Latency(CUDA preprocessing)	Optimization
yolov5s	41	28	31.7% $\downarrow$
yolov5lite	40	22	45% $\downarrow$
yolov6s	25	11	56% $\downarrow$
yolov7	47	32	31.9% $\downarrow$
yolov7_e2e	27	16	40.7% $\downarrow$

[wang-xinyu]

This CUDA preprocessing for YOLO is using warp affine method to do resizing, which is slightly different from cv::resize().
Hence the mAP is slightly different.
Below mAP(IoU=0.50:0.95 | area=all) results were tested on coco_val_2017, 5000 images, with TensorRT model.

Model	mAP(CPU preprocessing)	mAP(CUDA preprocessing)
yolov5s	0.372	0.368
yolov6s	0.424	0.418
yolov7	0.514	0.498

[jiangjiajun]

Windows的兼容下个PR再补充上