Data

Tools

Indexes

Applications

Experiments

Open Source Science

leap

Source code for implementing DPU, HDMI and IE on ZCU104 in PYNQ environment

https://github.com/jjsuperpower/leap

Science Score: 44.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ○
    Academic publication links
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (9.7%) to scientific vocabulary

Keywords

dpu-pynq pynq zcu104
Last synced: 7 months ago · JSON representation ·

Repository

Source code for implementing DPU, HDMI and IE on ZCU104 in PYNQ environment

Basic Info
  • Host: GitHub
  • Owner: jjsuperpower
  • License: mit
  • Language: Python
  • Default Branch: main
  • Homepage:
  • Size: 2.92 MB
Statistics
  • Stars: 5
  • Watchers: 1
  • Forks: 0
  • Open Issues: 1
  • Releases: 1
Topics
dpu-pynq pynq zcu104
Created over 2 years ago · Last pushed over 2 years ago
Metadata Files
Readme License Citation

README.md

LEAP

Integration of reaL-time image Enhancement and AI inferencing engine in the PYNQ environment

Overview

LEAP provides a platform for integrating real-time image enhancement and AI inferencing engine in the PYNQ environment for Xilinx's MPSoC ZCU104 FPGA development board. The computation of image enhancement and AI inference in processed in the Program Logic (PL) why being controlled via Python scripts running in Linux. The platform is designed to allow for straight forward integration of new image enhancement IPs and AI models. As a proof-of-concept we use an image enhancement IP to enhance dark images before being inferenced on by CNN models in real-time.

LEAP System Overview

Leap Hardware Design

Infererencing

We provided multiple implementations of inferencing that can favor either throughput or latency

Inferencing Methods

Example

YOLOv3 on Dark COCO Image

YOLOv3 on Dark + Histogram Equalization on Dark COCO Image

Getting Started

This repository is composed of 3 main components: 1. Hardware Design - Details and implementation of the FPGA design 2. Custom Pynq Image - Details of modifying the PYNQ SD card image to increase the size of the Contiguous Memory Allocator (CMA). 3. Runtime Code - Details and implementation of the code that controls real-time HDMI in/out, AI inferencing, and image enhancement.

How To Cite

@misc{sanderson2023integration, title={System Integration of Xilinx DPU and HDMI for Real-Time inference in PYNQ Environment with Image Enhancement}, author={Jonathan Sanderson and Syed Rafay Hasan}, year={2023}, eprint={2312.09506}, archivePrefix={arXiv}, primaryClass={eess.IV} }

Owner

  • Name: -
  • Login: jjsuperpower
  • Kind: user

Coding SkyNet

Citation (CITATION.cff) 

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: "Sanderson"
  given-names: "Jonathan"
- family-names: "Hasan"
  given-names: "Syed"
title: "LEAP"
version: 1.0.0
doi: 
date-released: 2023-12-18
url: "https://github.com/jjsuperpower/LEAP"
preferred-citation:
  type: article
  authors:
  - family-names: "Sanderson"
    given-names: "Jonathan"
  - family-names: "Hasan"
    given-names: "Syed"
  doi:  10.48550/arXiv.2312.09509
  title: "System Integration of Xilinx DPU and HDMI for Real-Time inference in PYNQ Environment with Image Enhancement"
  year: 2023

GitHub Events

Total
  • Watch event: 3
  • Issue comment event: 1
Last Year
  • Watch event: 3
  • Issue comment event: 1

Dependencies

runtime/requirements.txt pypi
  • Pillow >=9.0.1
  • matplotlib >=3.5.1
  • numpy >=1.21.5
  • pycocotools >=2.0.7
  • pynq >=3.0.1
  • pynq_dpu >=2.5.1
  • scipy >=1.8.0