https://github.com/qengineering/yolofastest-ncnn-jetson-nano
YoloFastestV2 for a Jetson Nano
https://github.com/qengineering/yolofastest-ncnn-jetson-nano
aarch64 cpp deep-learning jetson-nano ncnn ncnn-model yolofastest yolofastest-v2
Last synced: 6 months ago
JSON representation
YoloFastestV2 for a Jetson Nano
- Host: GitHub
- URL: https://github.com/qengineering/yolofastest-ncnn-jetson-nano
- Owner: Qengineering
- License: bsd-3-clause
- Created: 2022-01-11T14:16:13.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2023-09-30T11:19:07.000Z (about 2 years ago)
- Last Synced: 2025-03-27T09:23:08.869Z (7 months ago)
- Topics: aarch64, cpp, deep-learning, jetson-nano, ncnn, ncnn-model, yolofastest, yolofastest-v2
- Language: C++
- Homepage: https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html
- Size: 19 MB
- Stars: 15
- Watchers: 1
- Forks: 1
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# YoloFastest Jetson-Nano
## YoloFastest V2 with the ncnn framework.
[](https://opensource.org/licenses/BSD-3-Clause)
A truly impressive YOLO family member. As long as the images are not too large and/or the objects are too small, very high frame rates are achieved with more than acceptable accuracy. Thanks [dog-qiuqiu](https://github.com/dog-qiuqiu/Yolo-FastestV2) for all the hard work.
Special adapt for a Jetson Nano, see [Q-engineering deep learning examples](https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html)
------------
## Benchmark.
| Model | size | mAP | Jetson Nano | RPi 4 1950 | RPi 5 2900 | Rock 5 |
| ------------- | :-----: | :-----: | :-------------: | :-------------: | :-----: | :-----: |
| [NanoDet](https://github.com/Qengineering/NanoDet-ncnn-Jetson-Nano) | 320x320 | 20.6 | 26.2 FPS | 13.0 FPS | 43.2 FPS |36.0 FPS |
| [NanoDet Plus](https://github.com/Qengineering/NanoDetPlus-ncnn-Jetson-Nano) | 416x416 | 30.4 | 18.5 FPS | 5.0 FPS | 30.0 FPS | 24.9 FPS |
| [PP-PicoDet](https://github.com/Qengineering/PP-PicoDet-ncnn-Jetson-Nano) | 320x320 | 27.0 | 24.0 FPS | 7.5 FPS | 53.7 FPS | 46.7 FPS |
| [YoloFastestV2](https://github.com/Qengineering/YoloFastestV2-ncnn-Jetson-Nano) | 352x352 | 24.1 | 38.4 FPS | 18.8 FPS | 78.5 FPS | 65.4 FPS |
| [YoloV2](https://github.com/Qengineering/YoloV2-ncnn-Jetson-Nano) 20| 416x416 | 19.2 | 10.1 FPS | 3.0 FPS | 24.0 FPS | 20.0 FPS |
| [YoloV3](https://github.com/Qengineering/YoloV3-ncnn-Jetson-Nano) 20| 352x352 tiny | 16.6 | 17.7 FPS | 4.4 FPS | 18.1 FPS | 15.0 FPS |
| [YoloV4](https://github.com/Qengineering/YoloV4-ncnn-Jetson-Nano) | 416x416 tiny | 21.7 | 16.1 FPS | 3.4 FPS | 26.8 FPS | 22.4 FPS |
| [YoloV4](https://github.com/Qengineering/YoloV4-ncnn-Jetson-Nano) | 608x608 full | 45.3 | 1.3 FPS | 0.2 FPS | 1.82 FPS | 1.5 FPS |
| [YoloV5](https://github.com/Qengineering/YoloV5-ncnn-Jetson-Nano) | 640x640 small | 22.5 | 5.0 FPS | 1.6 FPS | 14.9 FPS | 12.5 FPS |
| [YoloV6](https://github.com/Qengineering/YoloV6-ncnn-Jetson-Nano) | 640x640 nano | 35.0 | 10.5 FPS | 2.7 FPS | 25.0 FPS | 20.8 FPS |
| [YoloV7](https://github.com/Qengineering/YoloV5-ncnn-Jetson-Nano) | 640x640 tiny | 38.7 | 8.5 FPS | 2.1 FPS | 21.5 FPS | 17.9 FPS |
| [YoloV8](https://github.com/Qengineering/YoloV8-ncnn-Jetson-Nano) | 640x640 nano | 37.3 | 14.5 FPS | 3.1 FPS | 20.0 FPS | 16.3 FPS |
| [YoloV8](https://github.com/Qengineering/YoloV8-ncnn-Jetson-Nano) | 640x640 small | 44.9 | 4.5 FPS | 1.47 FPS | 11.0 FPS | 9.2 FPS |
| [YoloX](https://github.com/Qengineering/YoloX-ncnn-Jetson-Nano) | 416x416 nano | 25.8 | 22.6 FPS | 7.0 FPS | 34.2 FPS | 28.5 FPS |
| [YoloX](https://github.com/Qengineering/YoloX-ncnn-Jetson-Nano) | 416x416 tiny | 32.8 | 11.35 FPS | 2.8 FPS | 21.8 FPS | 18.1 FPS |
| [YoloX](https://github.com/Qengineering/YoloX-ncnn-Jetson-Nano) | 640x640 small | 40.5 | 3.65 FPS | 0.9 FPS | 9.0 FPS | 7.5 FPS |
20 Recognize 20 objects (VOC) instead of 80 (COCO)
------------
## Dependencies.
To run the application, you have to:
- The Tencent ncnn framework installed. [Install ncnn](https://qengineering.eu/install-ncnn-on-jetson-nano.html)
- OpenCV 64-bit installed. [Install OpenCV 4.5](https://qengineering.eu/install-opencv-4.5-on-raspberry-64-os.html)
- Code::Blocks installed. (```$ 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/YoloFastestV2-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:
James.mp4
parking.jpg
parking_tiny.jpg
YoloFastestV2.cpb
mainFV2.cpp
yolo-fastestv2.cpp
yolo-fastestv2.h
yolo-fastestv2-opt.bin
yolo-fastestv2-opt.param
------------
## Running the app.
To run the application load the project file YoloFastestV2.cbp in Code::Blocks. More info or
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).
Many thanks to [dog-qiuqiu](https://github.com/dog-qiuqiu/Yolo-FastestV2)
------------
[](https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=CPZTM5BB3FCYL)