Go to the wiki to see full documentation
See it in action!
- http://lf2-classroom-monitor.us-east-2.elasticbeanstalk.com/
- https://classroom-helper-lf2.herokuapp.com/ (Heroku backup)
By placing a Jetson Nano with a camera in front of the classroom, you'll be able to monitor your classroom with a web dashboard.
The program on Jetson Nano will take a shot periodically, do face & emotion recognition, and record the result to database. The webapp hosted on AWS will then pick up the new data and update the dashboard. From the dashboard, you'll know how many students have come to your class, and the general vibe of the classroom. With the positivity vs. time graph, you'll even know if your joke works!
Relavant links
Quick demo
- Web: Flask & Vue
- Web hosting: AWS EC2 (with Elastic Beanstalk)
- Database: AWS DynamoDB
- Live Capuring: Jetson Nano
- Face Recognition: Jetson Nano
- Emotion Recognition: AWS Rekognition
- SURPRISED
- HAPPY
- CALM
- CONFUSED
- SAD
- ANGRY
- FEAR
- DISGUSTED
- UNKNOWN
There are two versions of frontend available. The default one is written in Vue, and the other is written in vanilla HTML/JavaScript/CSS + jQuery, rendered with Flask's template. You can switch to the latter one by setting VUE to false (or just remove it) in your environment variables.
There are many features:
Theme switcher
You can toggle the theme by clicking the moon button at the top right.
Time slider
You can easily change the scope of time to watch by dragging the slider at the top of the page.
Attendance
You can see the last recorded attendance as well as the attendance vs. time chart to see the attendance of your classroom over time.
Emotion
You can see the positivity of the classroom over time as well as the emotion cloud to get an idea of the general vibe of the classroom in the scope of time you selected.
Specific students
Apart from seeing the overall condition of the classroom, you can also head to each student's page to see their attendance record.
git clone https://github.io/dlccyes/LF2.git
Supply environment variables
cp .env_bak .env
Fill .env with the correct AWS credentials. (Never commit it.)
Install Python dependencies
pip3 install -r requirements.txt
Install Node dependencies
cd vue
npm install
cd vue
npm run build
python3 application.py
And you're done! Head to http://127.0.0.1:5000 to see the dashboard.
# Install protobuf compiler
$ sudo apt-get install protobuf-compiler
# Install buildtools
$ sudo apt-get install build-essential make
# Install grpc packages
$ pip3 install -r grpc_requirements.txt
Python dependencies
# Install boto3
$ pip3 install boto3
# Install dotenv
$ pip3 install python-dotenv
Rename .env_bak to .env in the project root and supply it with the correct credentials for this script to work.
AWS credentials
Install & configure AWS CLI or supply the credentials directly as follow
# Setup your host environment
$ cd ~
$ mkdir .aws && cd .aws
$ vim credentials
$ vim config
Add access key in the credentials file and region in the config file.
# Install the face_recognition package on Jetson Nano
$ sudo apt-get install python3-pip cmake libopenblas-dev liblapack-dev libjpeg-dev
$ git clone https://github.com/JetsonHacksNano/installSwapfile
$ ./installSwapfile/installSwapfile.sh
$ wget http://dlib.net/files/dlib-19.17.tar.bz2
$ tar jxvf dlib-19.17.tar.bz2
$ cd dlib-19.17
# Remove "forward_algo = forward_best_algo;" in dlib/cuda/cudnn_dlibapi.cpp
$ sudo python3 setup.py install
$ sudo pip3 install face_recognition
# Install Tensorflow on Jetson Nano
$ pip3 install Cython
$ pip3 install numpy
$ pip3 install https://dl.google.com/coral/python/tflite_runtime-2.1.0.post1-cp36-cp36m-linux_aarch64.whl
# Install evdev
$ pip3 install evdev
Change the event number to the that corresponding to the scanner in jetson-nano/scanner.py
Run the gRPC server on your Jetson Nano
$ python3 jetson-nano/server.py
Register with student ID & photo
sudo python3 jetson-nano/register.py
In-class monitor
python3 jetson-nano/inclass_monitor.py