emotion_detection_app

Deep learning based emotion detection for video conferencing applications

Steps to reproduce working version of app (FaceEMR):

1. clone the directory
2. Install android studio
3. In Android studio, go to open project and choose the given directory in pop up window.
4. Connect the device and build+run the app on device.

tflite file has been converted and added to the repo. (converted_model.tflite)

The current version of working app which captures the image and displays the emotion for captured image is using pb model (not tflite).

pb model is stored in src/java/assets folder of FaceEMR app.

TFLITE functionality for the current code needs to be added.

Dependencies:

Android studio (Other dependencies will be automatically downloaded through gradle when you load project in the Android studio).

Current isssues:

• Interfacing WebRTC with deep learning module.