# <span style="font-size: 50px;">🤾</span>CareerHack 2021 (ＧuavaDB)
[TOC]
## Contest
* [Official Website](https://www.microsoft.com/taiwan/campus/events/2021careerhack/)
* [Official Github](https://github.com/sha1painkiller/careerhack2021-workshop-aiot)
* [Convert model to device](https://github.com/sha1painkiller/careerhack2021-workshop-aiot/blob/main/tinyml-lab-persondetect.md)

## Reference
### Analysis
* [Deep Learning Based Systems Developed for Fall Detection: A Review](https://ieeexplore.ieee.org/document/9186685)
* [77] [A Study on the Application of Convolutional NeuralNetworks to Fall Detection Evaluated with MultiplePublic Datasets](https://www.mdpi.com/1424-8220/20/5/1466)
* [Fall detection in older adults with mobile IoT devices and machine learning in the cloud and on the edge](https://www.sciencedirect.com/science/article/pii/S0020025520304886)
* [Triaxial Accelerometer-Based Falls and Activities of Daily Life Detection Using Machine Learning](https://www.mdpi.com/1424-8220/20/13/3777/htm)
* [Trends in Smart Manufacturing: Role of Humans and Industrial Robots in Smart Factories](https://www.researchgate.net/publication/340740697_Trends_in_Smart_Manufacturing_Role_of_Humans_and_Industrial_Robots_in_Smart_Factories)
* [The Importance of Human Touch in Smart Manufacturing](https://altizon.com/importance-human-touch-smart-manufacturing/)
* [Industrial, Factory & Construction Accident Statistics](https://www.trantercleere.co.uk/accident-at-work/factory-accident-claims/industrial-factory-construction-accident-statistics)
* [Manufacturing statistics in Great Britain, 2020](https://www.hse.gov.uk/Statistics/industry/manufacturing.pdf)

### Methods
* [1D Convolutional Neural Network Models for Human Activity Recognition](https://machinelearningmastery.com/cnn-models-for-human-activity-recognition-time-series-classification/)
* [Fight-or-flight Response](https://en.wikipedia.org/wiki/Fight-or-flight_response)
* [What Happens to Your Body During the Fight or Flight Response?](https://health.clevelandclinic.org/what-happens-to-your-body-during-the-fight-or-flight-response/)
* [What Is General Adaptation Syndrome?](https://www.healthline.com/health/general-adaptation-syndrome)

## Resource
### Official
* [TSMC Workshop Preparation Hands-on](https://docdro.id/ZPPUBgI)
* [20210119 AIoT Workshop](https://docdro.id/uVfecmS)
* [Example Avnetdev Code](https://github.com/sha1painkiller/azureiot-lab-avnetdevkit)
### Datasets
* [A dataset build using wearable inertial measurement and ECG sensors for activity recognition, fall detection and basic heart anomaly detection system](https://www.sciencedirect.com/science/article/pii/S2352340919310728)
* [559 Data Sets](https://archive.ics.uci.edu/ml/datasets.php)
* [Horse Colic Data Set](https://archive.ics.uci.edu/ml/datasets/Horse+Colic)
* [Localization Data for Person Activity Data Set](https://archive.ics.uci.edu/ml/datasets/Localization+Data+for+Person+Activity)
* [EMG Physical Action Data Set Data Set](https://archive.ics.uci.edu/ml/datasets/EMG+Physical+Action+Data+Set)
* [Vicon Physical Action Data Set Data Set](https://archive.ics.uci.edu/ml/datasets/Vicon+Physical+Action+Data+Set)
* [Daily and Sports Activities Data Set](https://archive.ics.uci.edu/ml/datasets/Daily+and+Sports+Activities)

## Avnet Azure Sphere MT3620 Starter Kit
.jpg)

## ESP32
### Install
[Install ESP32 on Arduino IDE](http://pizgchen.blogspot.com/2019/08/esp32-arduino-ide-esp32.html)
### [Datasheet](https://components101.com/microcontrollers/esp32-devkitc)
### 高感度聲音傳感器

void setup() {
  Serial.begin(115200);
  pinMode(13,INPUT);
}
void loop() {
  int snd = analogRead(13);
  Serial.println(snd);
  delay(100);
}

* [Audio Frequency Detector](https://create.arduino.cc/projecthub/lbf20012001/audio-frequency-detector-617856)
### Max30100 心率感測器
[Tutorial](http://www.esp32learning.com/code/esp32-and-max30100-heart-rate-monitor-sensor.php)

## News
[](https://www.ettoday.net/news/20190412/1421246.htm)
[](https://tw.appledaily.com/local/20210113/MDLGMIJPBZHFPMXZJ2NUONJGN4/)
[](https://www.chinatimes.com/realtimenews/20210111002682-260402?chdtv)
[](https://forum.gamer.com.tw/C.php?bsn=60561&snA=16533)

## Presentation
### Flow
* 動機 -> 工廠意外頻繁(david數據工廠意外人數) -> 員工傷害很大and工廠損失 -> 智慧工廠部屬及維修時仍需要人力 -> 當自動化程度高，將造成意外發生時不會立即被發現 -> 而現今大部分停止機制在設備上 -> 速度不夠，傷害已造成 -> 全新思維 -> 將偵測的部分從設備轉到人身上 -> 透過危險發生當下生理產生的變化偵測危險停止周邊設備(厲害動畫)

因此我們開始上網看了很多相關文獻，發現在生理學及心理學中有個名叫 fight or flight 主要是再說 -> 目前用呼吸變化作為判斷依據(https://en.wikipedia.org/wiki/Fight-or-flight_response)

收集呼吸頻率方法 -> 頻率分析證明吸及吐 -> Data收集及Label方式 -> ML搭建

實作架構圖

Demo

### Resource

心臟跳動頻率提升
呼吸頻率提升
臉、皮膚泛紅
抑制胃和上消化道的作用，使消化減慢或停止
人體的許多血管會收縮
擴張肌肉血管
抑制淚腺（負責淚液產生）和流涎
瞳孔放大
抑制脊髓反射
發抖