Merge pull request #5 from aiotads/main

Updated PoC and added case study to enhance user experience
InnoIPA · Jul 7, 2023 · 2e61d0b · 2e61d0b
2 parents 0ab9a93 + 3f3dfc8
commit 2e61d0b
Show file tree

Hide file tree

Showing 24 changed files with 137 additions and 72 deletions.
diff --git a/Readme.md b/Readme.md
@@ -16,31 +16,39 @@ Copyright 2022 AIoT Dev. Dep., IPA
 
 # EXMU-X261
 
-- [EXMU-X261](#exmu-x261)
 - [See What's New](#see-whats-new)
 - [Introduction](#introduction)
-- [Table of contents](#table-of-contents)
+- [Getting Started](#getting-started)
 - [How to open an issue and sync repo on github?](#how-to-open-an-issue-and-sync-repo-on-github)
 - [License](#license)
 
 # See What's New
-- [Release Notes](./docs/release.md)
-- [Dependencies Version](./docs/release.md)
+- [Changelogs](./docs/changelogs.md)
 
 # Introduction
 We have created this repository to provide a quick and easy way to get started with our product. Here are some tips on how to learn it. Additionally, if you have any ideas regarding our content, you can open an issue on GitHub. Sync fork will bring up the newest content from the upstream repository.
 
-# Table of contents
- - [Getting Started](./tocs/0.Getting-Started/Getting-Started.md)
+# Getting Started
+ - [Installation and Setup](./tocs/0.Getting-Started/Getting-Started.md)
   - [Hardware](./tocs/1.Hardware/hardware.md)
   - [Software](./tocs/2.Software/)
     - [dpu-sc](./tocs/2.Software/dpu-sc.md)
     - [stesting-sc](./tocs/2.Software/stesting-sc.md)
-    - [VVAS](./tocs/2.Software/VVAS.md)
-  - [POC](./tocs/3.POC/)
-    - [Defect Detection](./tocs/3.POC/Defect-Detection.md)
+    - [VVAS(Vitis Video Analytics SDK)](./tocs/2.Software/VVAS.md)
+  - [PoC](./tocs/3.POC/)
     - [VVAS Demo](./tocs/3.POC/VVAS-Demo.md)
-  - [FAQ](./tocs/4.FAQ/FAQ.md)
+    - [ANPR](./tocs/3.POC/ANPR.md)
+  - [Case study](./tocs/5.Case-study/case-study.md)
+    - [Powered by AMD](https://www.xilinx.com/video/corporate/innodisk-machine-vision-solution-kit-powered-by-amd-kria-som.html)
+      - [Innodisk’s Machine Vision Solution Kit Powered by AMD’s Kria SOM](https://www.xilinx.com/content/dam/xilinx/publications/powered-by-xilinx/innodisk-case-study.pdf)
+    - [Defects detection solutions](./tocs/5.Case-study/Defect-Detection.md)
+        - **USB Pendrive assembly**
+        - **Screw defect**
+        - **Pneumonia (X-Ray)**
+    - [用于缺陷检测的自动化边缘 AI 解决方案](https://china.xilinx.com/video/events/cns_xtd_AI_solution.html)
+    - [宜鼎国际最新肺部X光片偵測 自动化边缘 AI 解决方案加速医疗诊断速度](https://www.myinnodisk.cn/newsDetail/1363.html)
+    - [Innodisk Corporation(Platform) x Neovision(AI Model)](https://www.linkedin.com/posts/neovisionsas_intelligenceartificielle-industrie-activity-7039869643737456640-52r8?utm_source=share&utm_medium=member_desktop)
+      - **Gesture Recognition**
 
   <br />
   <div align="center"><img width="80%" height="50%" src="./EXMU-X261.png"></div>

diff --git a/docs/changelogs.md b/docs/changelogs.md
@@ -0,0 +1,24 @@
+<!--
+ Copyright (c) 2022 Innodisk Crop.
+ 
+ This software is released under the MIT License.
+ https://opensource.org/licenses/MIT
+-->
+
+# Changelogs
+
+Type | desc. | version | Autor
+--- | --- | --- | ---
+Feat | Introduction of EXMU-X261 hardware<br/> Introduction of EXMU-X261 BSP v0.0.3 LTS<br/> Introduction of EXMU-X261 software application | 0.0.1 | Ju (hueiru_chen@innodisk.com)
+doc | Add workflow of flashing qspi<br/> Update workflow of flashing emmc<br/> Update A2 carrier board info<br/> Update rpm discription<br/> Update "What is BSP"<br/> Modify FAQ and dependency versions<br/> Fix typos<br/> | 0.0.2 | Ju (hueiru_chen@innodisk.com)
+doc | Update BSP v1.0.4 for A3 carrier board | 0.0.3 | Ju (hueiru_chen@innodisk.com)
+doc | 1. Modify release note to changelog <br/> 2. add case study section <br/> 3. Modify wording <br/> 4. add ANPR.md <br/> 5. Rearrange defect detection| 0.0.4 | AH (allen_huang@innodisk.com)
+
+# Dependencies Version
+
+BSP | Vitis-AI | VVAS|stesting|
+|:---:|:---:|:---:|:---:|
+|1.0.4|1.4|1.0|0.0.4|
+
+- The dpu-sc would not be preload in BSP.
+- The VVAS not be support before BSP version v0.0.3. Would be support start from v1.0.4 version.
diff --git a/docs/release.md b/docs/release.md
diff --git a/tocs/0.Getting-Started/Getting-Started.md b/tocs/0.Getting-Started/Getting-Started.md
@@ -4,8 +4,7 @@
  This software is released under the MIT License.
  https://opensource.org/licenses/MIT
 -->
-# Table of Contents
-- [Table of Contents](#table-of-contents)
+# TOC
 - [What you will need](#what-you-will-need)
   - [Hardware](#hardware)
   - [Software](#software)

diff --git a/tocs/1.Hardware/hardware.md b/tocs/1.Hardware/hardware.md
@@ -5,6 +5,7 @@
  https://opensource.org/licenses/MIT
 -->
 
+# TOC
 - [SOM Board Overview](#som-board-overview)
 - [Devices on SOM](#devices-on-som)
   - [QSPI Flash](#qspi-flash)

diff --git a/tocs/2.Software/VVAS.md b/tocs/2.Software/VVAS.md
@@ -4,7 +4,7 @@
  This software is released under the MIT License.
  https://opensource.org/licenses/MIT
 -->
-
+# TOC
 - [Overview](#overview)
 - [How to Install](#how-to-install)
   - [Install by RPM](#install-by-rpm)

diff --git a/tocs/2.Software/dpu-sc.md b/tocs/2.Software/dpu-sc.md
@@ -4,6 +4,9 @@
  This software is released under the MIT License.
  https://opensource.org/licenses/MIT
 -->
+# TOC
+- [Overview](#overview)
+- [Download](#download)
 
 # Overview
 [dpu-sc](https://github.com/InnoIPA/dpu-sc)(DPU sample code) is a public repo. DPU-SC is a python sample code. It shows how to run AI inference on our system. Also, user can replace customized model(.xmodel) to use it. 

diff --git a/tocs/2.Software/stesting-sc.md b/tocs/2.Software/stesting-sc.md
@@ -4,7 +4,7 @@
  This software is released under the MIT License.
  https://opensource.org/licenses/MIT
 -->
-
+# TOC
 - [Overview](#overview)
 - [Download](#download)
 

diff --git a/tocs/3.POC/ANPR.md b/tocs/3.POC/ANPR.md
@@ -0,0 +1,18 @@
+<!--
+ Copyright (c) 2022 Innodisk crop.
+ 
+ This software is released under the MIT License.
+ https://opensource.org/licenses/MIT
+-->
+# TOC
+- [Overview](#overview)
+- [Demo](#Demo)
+
+# Overview
+Automatic number-plate recognition is a technology that uses optical character recognition on images to read vehicle registration plates to create vehicle location data. Source: [Wiki](https://en.wikipedia.org/wiki/Automatic_number-plate_recognition)
+
+LPRNet: License Plate Recognition via Deep Neural Networks: end-to-end method for Automatic License Plate Recognition without preliminary character segmentation. [Paper](https://arxiv.org/abs/1806.10447)
+
+# Demo
+We have implemented the demo in [dpu-sc](../2.Software/dpu-sc.md).
+![LPR](./fig/lpr.gif)
diff --git a/tocs/3.POC/Defect-Detection.md b/tocs/3.POC/Defect-Detection.md
diff --git a/tocs/3.POC/VVAS-Demo.md b/tocs/3.POC/VVAS-Demo.md
@@ -5,6 +5,7 @@
  https://opensource.org/licenses/MIT
 -->
 
+# TOC
 - [Overview](#overview)
 - [How to Install](#how-to-install)
   - [VVAS](#vvas)
@@ -14,6 +15,8 @@
   - [Innodisk demo](#innodisk-demo)
     - [Install by RPM](#install-by-rpm-1)
 - [How to run](#how-to-run)
+- [DEMO](#demo)
+  - [Computex taipei, 2023](#computex-taipei-2023)
 - [Reference](#reference)
 
 # Overview
@@ -130,6 +133,10 @@ rpm -ivh --force vvas_demo-0.1-1.aarch64.rpm
    ![vvas_demo](fig/vvas_demo.jpg)
 
 
+# DEMO
+## Computex taipei, 2023
+![computex](fig/vvas_computex_2023.gif)
+
 # Reference
 
 [Multichannel ML](https://xilinx.github.io/VVAS/main/build/html/docs/Embedded/Tutorials/MultiChannelML.html)

diff --git a/tocs/3.POC/fig/lpr.gif b/tocs/3.POC/fig/lpr.gif
diff --git a/tocs/3.POC/fig/vvas_computex_2023.gif b/tocs/3.POC/fig/vvas_computex_2023.gif
diff --git a/tocs/4.FAQ/FAQ.md b/tocs/4.FAQ/FAQ.md
@@ -7,7 +7,6 @@
 
 
 # FAQ
-- [FAQ](#faq)
 - [What is BSP?](#what-is-bsp)
 - [How to flash eMMC on X261?](#how-to-flash-emmc-on-x261)
 - [How to update QSPI FW?](#how-to-update-qspi-fw)

diff --git a/tocs/5.Case-study/Defect-Detection.md b/tocs/5.Case-study/Defect-Detection.md
@@ -0,0 +1,57 @@
+<!--
+ Copyright (c) 2022 Innodisk crop.
+ 
+ This software is released under the MIT License.
+ https://opensource.org/licenses/MIT
+-->
+# TOC
+- [Intorduction](#intorduction)
+- [Defect Detection Solutions](#defect-detection-solutions)
+  - [USB pendrive Defect Detection](#usb-pendrive-defect-detection)
+  - [Screw Defect Detection](#screw-defect-detection)
+  - [Pneumonia Detection](#pneumonia-detection)
+
+# Intorduction
+We create the defect detection solutions on EXMU-X261. When camera catch the image, the image will input to EXMU-X261. The EXMU-X261 will preprocess the image. After all, the DPU will inference image and output result with bounding box. The result will shows on screen.  
+
+![DD-flow](./fig/DD-flow.png)
+
+|Solution| Model | Input Size | AVG. FPS on X261 |
+|:------:| :-------------: |:----: |:---:|
+| Object detection | YOLOv4-tiny | 320x320  | ~35 FPS |
+
+`In the following section we have presented 3 case studies for you with object detection solution.`
+
+# Defect Detection Solutions 
+<details>
+<summary> USB pendrive Defect Detection</summary>  
+
+This solution can be used in factory to detect the logo or the assembly of components error.
+
+PASS
+![DD-OK](./fig/DD_predict_result_true.png)
+
+NG
+![DD-False](./fig/DD_predict_result_false.png)
+
+Live demo  
+![DD-USB](./fig/DD-USB.gif)
+
+</details>
+</br>
+<details>
+<summary>Screw Defect Detection</summary>  
+
+This solution can be used in factory to detect the screw defect.
+
+![screw_predict_result](./fig/DD-screw.gif)
+</details>
+</br>
+<details>
+<summary>Pneumonia Detection</summary>
+
+[COVID-19 Sniper: An advanced diagnostic taool for pneumonia](https://www.hackster.io/covid-19-sniper/covid-19-sniper-an-advanced-diagnostic-tool-for-pneumonia-70be48)  
+
+![oty](./fig/DD-oty.gif)
+</details>
+
diff --git a/tocs/5.Case-study/case-study.md b/tocs/5.Case-study/case-study.md
@@ -0,0 +1,6 @@
+<!--
+ Copyright (c) 2022 Innodisk crop.
+ 
+ This software is released under the MIT License.
+ https://opensource.org/licenses/MIT
+-->
diff --git a/tocs/3.POC/fig/DD-USB.gif → tocs/5.Case-study/fig/DD-USB.gif b/tocs/3.POC/fig/DD-USB.gif → tocs/5.Case-study/fig/DD-USB.gif
diff --git a/tocs/3.POC/fig/DD-flow.drawio → tocs/5.Case-study/fig/DD-flow.drawio b/tocs/3.POC/fig/DD-flow.drawio → tocs/5.Case-study/fig/DD-flow.drawio
diff --git a/tocs/3.POC/fig/DD-flow.png → tocs/5.Case-study/fig/DD-flow.png b/tocs/3.POC/fig/DD-flow.png → tocs/5.Case-study/fig/DD-flow.png
diff --git a/tocs/5.Case-study/fig/DD-oty.gif b/tocs/5.Case-study/fig/DD-oty.gif
diff --git a/tocs/5.Case-study/fig/DD-screw.gif b/tocs/5.Case-study/fig/DD-screw.gif
diff --git a/tocs/3.POC/fig/DD_predict_result_false.png → ...ase-study/fig/DD_predict_result_false.png b/tocs/3.POC/fig/DD_predict_result_false.png → ...ase-study/fig/DD_predict_result_false.png
diff --git a/tocs/3.POC/fig/DD_predict_result_true.png → ...Case-study/fig/DD_predict_result_true.png b/tocs/3.POC/fig/DD_predict_result_true.png → ...Case-study/fig/DD_predict_result_true.png
diff --git a/tocs/3.POC/fig/screw_predict_result.png → ...ase-study/fig/DD_screw_predict_result.png b/tocs/3.POC/fig/screw_predict_result.png → ...ase-study/fig/DD_screw_predict_result.png