BabyReady uses a Convolutional Neural Network to predict the reason a baby is crying with a mp3 recording as input.

BabyReady was developed in 2018 by Ang Kang Rong Roy and Kyle Zheng Ching Chan with help from Tang Yetong. It won the champion and healthtech award at Splash Awards 2018 and first runner-up in CodeXtremeApps 2018.

1. Run the BabyReady android app and place device beside baby.
2. When baby cries, the cry is recorded and uploaded to the online server which predicts the reason
3. The reason is sent to the android device through an online database and the device will act based on the predicted reason. If the baby is scared or lonely, Baby Shark is played. If the baby is tired, a lullaby is played.

You must install both the server and android app.

1. Download the entire BabyReady repository.
2. Create your own Firebase project and download the google-servies.json file. Add it to both BabyReady/BabyReady - Python/auth.json and BabyReady/BabyReady - Android/app/google-services.json. Create an environmental variable on your computer GOOGLE_APPLICATION_CREDENTIALS and enter as value the file location of BabyReady/BabyReady - Python/auth.json.
3. pip install -r requirements.txt.
4. Build the apk from the Android project and install in android device.

1. Add data to BabyReady/BabyReady - Python/data/
2. Run BabyReady/BabyReady - Python/train_set.py
3. Run BabyReady/BabyReady - Python/train_model.py

A sample model has already been created in BabyReady/BabyReady - Python/output/. You may continue with this if you do not want to create a new model yourself.

Start server by running BabyReady/BabyReady - Python/try_firebase.py on your computer Start client by running android apk built from Android Project in BabyReady/BabyReady - Android/

Alternatively, just run BabyReady/BabyReady - Python/BabyCryRecording.py on your computer

Codes for each reason for baby cry: bp - belly pain

bu - burp

ch - temperature

dc - discomfort

hu - hungry

lo - lonely

sc - scared

ti - tired