YOLOv10 Example

Usage

Make sure you have downloaded the data files first for the examples. You only need to do this once for all examples.

cd example/
git clone --depth=1 https://github.com/swdee/go-rknnlite-data.git data

Run the YOLOv10 example on rk3588 or replace with your Platform model.

cd example/yolov10
go run yolov10.go -p rk3588

This will result in the output of:

Driver Version: 0.9.6, API Version: 2.3.0 (c949ad889d@2024-11-07T11:35:33)
Model Input Number: 1, Ouput Number: 6
Input tensors:
  index=0, name=images, n_dims=4, dims=[1, 640, 640, 3], n_elems=1228800, size=1228800, fmt=NHWC, type=INT8, qnt_type=AFFINE, zp=-128, scale=0.003922
Output tensors:
  index=0, name=485, n_dims=4, dims=[1, 64, 80, 80], n_elems=409600, size=409600, fmt=NCHW, type=INT8, qnt_type=AFFINE, zp=-62, scale=0.086849
  index=1, name=499, n_dims=4, dims=[1, 80, 80, 80], n_elems=512000, size=512000, fmt=NCHW, type=INT8, qnt_type=AFFINE, zp=-128, scale=0.002931
  index=2, name=506, n_dims=4, dims=[1, 64, 40, 40], n_elems=102400, size=102400, fmt=NCHW, type=INT8, qnt_type=AFFINE, zp=-55, scale=0.072764
  index=3, name=520, n_dims=4, dims=[1, 80, 40, 40], n_elems=128000, size=128000, fmt=NCHW, type=INT8, qnt_type=AFFINE, zp=-128, scale=0.003646
  index=4, name=527, n_dims=4, dims=[1, 64, 20, 20], n_elems=25600, size=25600, fmt=NCHW, type=INT8, qnt_type=AFFINE, zp=-68, scale=0.058066
  index=5, name=541, n_dims=4, dims=[1, 80, 20, 20], n_elems=32000, size=32000, fmt=NCHW, type=INT8, qnt_type=AFFINE, zp=-128, scale=0.003894
bus @ (92 136 555 436) 0.954108
person @ (110 234 226 536) 0.911271
person @ (212 240 285 509) 0.872328
person @ (477 233 559 521) 0.825596
person @ (80 330 123 514) 0.488516
Model first run speed: inference=47.048717ms, post processing=1.271934ms, rendering=713.403µs, total time=49.034054ms
Saved object detection result to ../data/bus-yolov10-out.jpg
Benchmark time=4.091851863s, count=100, average total time=40.918518ms
done

The saved JPG image with object detection markers.

To use your own RKNN compiled model and images.

go run yolov10.go -m <RKNN model file> -i <image file> -l <labels txt file> -o <output jpg file> -p <platform>

The labels file should be a text file containing the labels the Model was trained on. It should have one label per line.

See the help for command line parameters.

$ go run yolov10.go --help

Usage of /tmp/go-build859033258/b001/exe/yolov10:
  -i string
        Image file to run object detection on (default "../data/bus.jpg")
  -l string
        Text file containing model labels (default "../data/coco_80_labels_list.txt")
  -m string
        RKNN compiled YOLO model file (default "../data/models/rk3588/yolov10s-rk3588.rknn")
  -o string
        The output JPG file with object detection markers (default "../data/bus-yolov10-out.jpg")
  -p string
        Rockchip CPU Model number [rk3562|rk3566|rk3568|rk3576|rk3582|rk3582|rk3588] (default "rk3588")

Docker

To run the YOLOv10 example using the prebuilt docker image, make sure the data files have been downloaded first, then run.

# from project root directory

docker run --rm \
  --device /dev/dri:/dev/dri \
  -v "$(pwd):/go/src/app" \
  -v "$(pwd)/example/data:/go/src/data" \
  -v "/usr/include/rknn_api.h:/usr/include/rknn_api.h" \
  -v "/usr/lib/librknnrt.so:/usr/lib/librknnrt.so" \
  -w /go/src/app \
  swdee/go-rknnlite:latest \
  go run ./example/yolov10/yolov10.go -p rk3588

Proprietary Models

The example YOLOv10 model used has been trained on the COCO dataset so makes use of the default Post Processor setup. If you have trained your own Model and have set specific Classes or want to use alternative Box and NMS Threshold values, then initialize the postprocess.NewYOLOv10 with your own YOLOv10Params.

In the file postprocess/yolov10.go see function YOLOv10COCOParams for how to configure your own custom parameters.

Benchmarks

The following table shows a comparison of the benchmark results across the three distinct platforms.

Platform	Execution Time	Average Inference Time Per Image
rk3588	4.09s	40.91ms
rk3576	3.52s	35.28ms
rk3566	11.79s	117.92ms

Note that these examples are only using a single NPU core to run inference on. The results would be different when running a Pool of models using all NPU cores available. Secondly the Rock 4D (rk3576) has DDR5 memory versus the Rock 5B (rk3588) with slower DDR4 memory.

Background

This YOLOv10 example is a Go conversion of the C API example.