Ecosyste.ms: Awesome

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

Awesome Lists | Featured Topics | Projects

https://github.com/LukiBa/zybo_yolo

YOLO example implementation using Intuitus CNN accelerator on ZYBO ZYNQ-7000 FPGA board
https://github.com/LukiBa/zybo_yolo

Last synced: 3 months ago
JSON representation

YOLO example implementation using Intuitus CNN accelerator on ZYBO ZYNQ-7000 FPGA board

Host: GitHub
URL: https://github.com/LukiBa/zybo_yolo
Owner: LukiBa
License: mit
Created: 2021-02-17T08:37:24.000Z (about 4 years ago)
Default Branch: master
Last Pushed: 2021-07-30T09:18:19.000Z (over 3 years ago)
Last Synced: 2024-08-02T01:20:57.544Z (7 months ago)
Language: Python
Size: 137 MB
Stars: 16
Watchers: 3
Forks: 2
Open Issues: 0
Metadata Files:
- Readme: README.md
- License: LICENSE

Awesome Lists containing this project

awesome-yolo-object-detection - LukiBa/zybo_yolo - 7000 FPGA board. (Lighter and Deployment Frameworks)

README

        # Zybo Yolo

![GitHub](https://img.shields.io/github/license/LukiBa/zybo_yolo)  

YOLOv3 tiny example implementation for Intuitus CNN accelerator using a ZYBO-Z7-20 FPGA board. 

## Features

- [x] Capture image using v4l2 camera 

- [x] Plot results to framebuffer 

- [x] Run Yolov3 tiny inference on FPGA 

- [x] Inference time downto 82 ms

- [x] Low power consumption due to FPGA implementation

## Run test application:

````sh

python zybo_yolo/test-yolov3-tiny.py --use_cam --use_fb

````

## Hardware Requirements

- Zybo-Z7-20 Board 

- microSD card >=8GB 

- Optional:

    - PCAM-5C 

    - HDMI Display 

## Setup Hardware

1. SD card: Format first Partition (300-500MB) to FAT and the second partition to EXT4

2. Copy BOOT.BIN and image.ub to the first partition 

3. Download linaro-developer and extract it to the second partition 

4. Copy intuitus.ko to extracted rootfs.

5. Connect ethernet, pcam, display, keyboard and power to Zybo and insert the SD card

6. Boot up the system

7. Update the OS using ``sudo apt-get upgrade && apt-get update``

8. Install v4l2, cmake, opencv and swig

9. Install PyEnv

10. Install python 3.6.10 and activate it

11. Install numpy, opencv-python, elevate, pathlib, wget and argparse using pip 

12. Clone 

13. Install the interface using ``pip install -e .`` 

14. Clone 

15. cd to intuitus.ko location and insert device driver using ``sudo insmod intuitus.ko``

16. Run test application using ``python zybo_yolo/test-yolov3-tiny.py --use_cam --use_fb``

## Results ^[1] ^[2] 

| Mem Clk [MHz] | PE Clk [MHz] | P_idle [W] | P_exec [W] | inference time [ms] |

| ------ | ------ | ------ | ------ | ------ |

| 100 | 100 | 2.73 | 3.12 | 112 |

| 100 | 175 | 3.32 | 3.85 | 91 |

| 100 | 200 | 3.60 | 4.10 | 89 |

| 125 | 200 | 3.57 | 4.10 | 86 |

| 125 | 225 | 3.83 | 4.35 | 84 |

| 125 | 250 | 4.02 | 4.57 | 83 |

[1] The average time of 1000 calls of the Network is used as a measure for inference time (input size: 384x384).  

[2] The power is the power consumption of the board without a connected PCAM and display.

![Example output](./readme_detect.png)  

## Related Projects: 

| Project | link |

| ------ | ------ |

| Intuitus Model Converter |  |

| Intuitus Interface | https://github.com/LukiBa/Intuitus-intf.git |

| Intuitus device driver | link to kernel module comming soon (contact author for sources) |

| Vivado Yolov3-tiny example project | https://github.com/LukiBa/zybo_yolo_vivado.git |

| Intuitus FPGA IP | encrypted trial version comming soon (contact author) |

## Author

Lukas Baischer   

[email protected]