Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/Qengineering/YoloV10-ncnn-Raspberry-Pi-4

YoloV10 for a bare Raspberry Pi 4 or 5
https://github.com/Qengineering/YoloV10-ncnn-Raspberry-Pi-4

aarch64 cpp ncnn ncnn-model raspberry-pi raspberry-pi-4 raspberry-pi-5 yolov10

Last synced: about 2 months ago
JSON representation

YoloV10 for a bare Raspberry Pi 4 or 5

Host: GitHub
URL: https://github.com/Qengineering/YoloV10-ncnn-Raspberry-Pi-4
Owner: Qengineering
License: bsd-3-clause
Created: 2024-06-11T13:16:57.000Z (3 months ago)
Default Branch: main
Last Pushed: 2024-06-21T15:44:32.000Z (3 months ago)
Last Synced: 2024-07-08T06:47:01.524Z (2 months ago)
Topics: aarch64, cpp, ncnn, ncnn-model, raspberry-pi, raspberry-pi-4, raspberry-pi-5, yolov10
Language: C++
Homepage: https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html
Size: 4.84 MB
Stars: 16
Watchers: 1
Forks: 1
Open Issues: 0
Metadata Files:
- Readme: README.md
- License: LICENSE

Awesome Lists containing this project

awesome-yolo-object-detection - Qengineering/YoloV10-ncnn-Raspberry-Pi-4 - ncnn-Raspberry-Pi-4?style=social"/> : YoloV10 for a bare Raspberry Pi 4 or 5. (Lighter and Deployment Frameworks)

README

        # YoloV10 Raspberry Pi 4 or 5

![output image]( https://qengineering.eu/github/YoloV10_parking.jpg )

## YoloV10 with the ncnn framework. 


[![License](https://img.shields.io/badge/License-BSD%203--Clause-blue.svg)](https://opensource.org/licenses/BSD-3-Clause)



Paper: https://arxiv.org/pdf/2405.14458 



Special made for a bare Raspberry Pi 4, see [Q-engineering deep learning examples](https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html)

------------

## Benchmark.

Numbers in **FPS** and reflect only the inference timing. Grabbing frames, post-processing and drawing are not taken into account.

| Model  | size | mAP | Jetson Nano | RPi 4 1950 | RPi 5 2900 | Rock 5 | RK3588¹
NPU | RK3566/68²
NPU | Nano
TensorRT | Orin
TensorRT |

| ------------- | :-----:  | :-----:  | :-------------:  | :-------------: | :-----: | :-----: | :-------------:  | :-------------: | :-----: | :-----: |

| [NanoDet](https://github.com/Qengineering/NanoDet-ncnn-Raspberry-Pi-4) | 320x320 | 20.6  |  26.2 | 13.0 | 43.2  |36.0  |||||

| [NanoDet Plus](https://github.com/Qengineering/NanoDetPlus-ncnn-Raspberry-Pi-4) | 416x416 | 30.4  |  18.5  | 5.0  | 30.0  | 24.9  |||||

| [PP-PicoDet](https://github.com/Qengineering/PP-PicoDet-ncnn-Raspberry-Pi-4) | 320x320 | 27.0  |  24.0 | 7.5 | 53.7 | 46.7 |||||

| [YoloFastestV2](https://github.com/Qengineering/YoloFastestV2-ncnn-Raspberry-Pi-4) | 352x352 | 24.1 |  38.4 | 18.8 | 78.5 | 65.4 | ||||

| [YoloV2](https://github.com/Qengineering/YoloV2-ncnn-Raspberry-Pi-4) ²⁰| 416x416 | 19.2 |  10.1 | 3.0 | 24.0 | 20.0 | ||||

| [YoloV3](https://github.com/Qengineering/YoloV3-ncnn-Raspberry-Pi-4) ²⁰| 352x352 tiny | 16.6 | 17.7 | 4.4 | 18.1 | 15.0 | ||||

| [YoloV4](https://github.com/Qengineering/YoloV4-ncnn-Raspberry-Pi-4) | 416x416 tiny | 21.7 | 16.1 | 3.4 | 17.5 | 22.4 | ||||

| [YoloV4](https://github.com/Qengineering/YoloV4-ncnn-Raspberry-Pi-4) | 608x608 full | 45.3 | 1.3 | 0.2 | 1.82 | 1.5 | ||||

| [YoloV5](https://github.com/Qengineering/YoloV5-ncnn-Raspberry-Pi-4) | 640x640 nano | 22.5 | 5.0 | 1.6 | 13.6 | 12.5 | 58.8 | 14.8 | 19.0 | 100 |

| [YoloV5](https://github.com/Qengineering/YoloV5-ncnn-Raspberry-Pi-4) | 640x640 small | 22.5 | 5.0 | 1.6 | 6.3 | 12.5 | 37.7 | 11.7 | 9.25 | 100 |

| [YoloV6](https://github.com/Qengineering/YoloV6-ncnn-Raspberry-Pi-4) | 640x640 nano | 35.0 | 10.5 | 2.7 | 15.8 | 20.8 | 63.0 | 18.0 |||

| [YoloV7](https://github.com/Qengineering/YoloV5-ncnn-Raspberry-Pi-4) | 640x640 tiny | 38.7 | 8.5 | 2.1 | 14.4 | 17.9 |  53.4 | 16.1 | 15.0 ||

| [YoloV8](https://github.com/Qengineering/YoloV8-ncnn-Raspberry-Pi-4) | 640x640 nano | 37.3 | 14.5 | 3.1 | 20.0 | 16.3 | 53.1 | 18.2 |||

| [YoloV8](https://github.com/Qengineering/YoloV8-ncnn-Raspberry-Pi-4) | 640x640 small | 44.9 | 4.5 | 1.47 | 11.0 | 9.2 | 28.5 | 8.9 |||

| [YoloV9](https://github.com/Qengineering/YoloV9-ncnn-Raspberry-Pi-4) | 640x640 comp | 53.0 | 1.2 | 0.28 | 1.5 | 1.2 | |||| 

| [YoloV10](https://github.com/Qengineering/YoloV10-ncnn-Raspberry-Pi-4) | 640x640 nano | 38.0 | | 1.66 | | | |||| 

| [YoloX](https://github.com/Qengineering/YoloX-ncnn-Raspberry-Pi-4) | 416x416 nano | 25.8 | 22.6 | 7.0 | 38.6 | 28.5 | ||||

| [YoloX](https://github.com/Qengineering/YoloX-ncnn-Raspberry-Pi-4) | 416x416 tiny | 32.8 | 11.35 | 2.8 | 17.2 | 18.1 | ||||

| [YoloX](https://github.com/Qengineering/YoloX-ncnn-Raspberry-Pi-4) | 640x640 small | 40.5 | 3.65 | 0.9 | 4.5 | 7.5 | 30.0 | 10.0 |||

¹ The Rock 5 and Orange Pi5 have the RK3588 on board.


² The Rock 3, Radxa Zero 3 and Orange Pi3B have the RK3566 on board.


²⁰ Recognize 20 objects (VOC) instead of 80 (COCO)

------------

## Opset.

YoloV10 use layer operations yet unknown to ncnn. To let YoloV10 run on an ncnn framework, [zhouweigogogo](https://github.com/zhouweigogogo) has modified the onnx file by replacing these layers. 

At the same time, he expanded the original output to three. Some features, like variable input size, aren't supported now. 

Until the ncnn framework fully supports onnx opset 13, we face these limitations.

------------

## Dependencies.

To run the application, you have to:

- A Raspberry Pi 4 or 5 with a 32 or 64-bit operating system. It can be the Raspberry 64-bit OS, or Ubuntu 18.04 / 20.04. [Install 64-bit OS](https://qengineering.eu/install-raspberry-64-os.html) 


- The Tencent ncnn framework installed. [Install ncnn](https://qengineering.eu/install-ncnn-on-raspberry-pi-4.html) 


- OpenCV 64-bit installed. [Install OpenCV 4.5](https://qengineering.eu/install-opencv-4.5-on-raspberry-64-os.html) 


- Code::Blocks installed (Optional). (```$ sudo apt-get install codeblocks```)

------------

## Installing the app.

To extract and run the network in Code::Blocks 


$ mkdir *MyDir* 


$ cd *MyDir* 


$ wget https://github.com/Qengineering/YoloV10-ncnn-Raspberry-Pi-4/archive/refs/heads/main.zip 


$ unzip -j master.zip 


Remove master.zip, LICENSE and README.md as they are no longer needed. 
 

$ rm master.zip 


$ rm LICENSE 


$ rm README.md 
 


Your *MyDir* folder must now look like this: 
 

parking.jpg 


busstop.jpg 


YoloV10.cpb 


yoloV10.cpp 


yolov10n.ncnn.bin 


yolov10n.ncnn.param 


------------

## Running the app.

You can use **Code::Blocks**.

- Load the project file *.cbp in Code::Blocks.

- Select _Release_, not Debug.

- Compile and run with F9.

- You can alter command line arguments with _Project -> Set programs arguments..._ 

Or use **Cmake**.

```

$ cd *MyDir*

$ mkdir build

$ cd build

$ cmake ..

$ make -j4

```

if you want to connect a camera to the app, follow the instructions at [Hands-On](https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html#HandsOn).


------------

### Thanks.

A more than special thanks to [***THU-MIG***](https://github.com/THU-MIG), who adapted the YoloV10 to onnx models.


And [***zhouweigogogo***](https://github.com/zhouweigogogo), who adapted the onnx model to the ncnn framework.



![output image]( https://qengineering.eu/github/YoloV10_busstop.jpg )

------------

[![paypal](https://qengineering.eu/images/TipJarSmall4.png)](https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=CPZTM5BB3FCYL)