Project description
This is a sample web app using of Affectiva Java SDK. Some of the codes here are based on Affectiva instructions. I nade some modifications to run it locally. In the future, our team will develop more features and intergrate this module into the main app.
- Use Chrome browser to install and run https://chrome.google.com/webstore/detail/web-server-for-chrome/ofhbbkphhbklhfoeikjpcbhemlocgigb
- In "Choose Folder" button, open your local repo that contains "index.html" file. Check "Automatically show index.html"
- "Stop" then "Start" the switch right under "Choose Folder" button
Having installed dependencies, run the prediction, passing trained model as well as a path to the file you want to evaluate. In my case, I developed the model in a conda environment.
Invocation is performed as follows: $ ./predict path/to/model.h5 path-to/file.mpg
example: $ ./predict evaluation/models/unseen-weights178.h5 id2_vcd_swwp2s.mpg
If the dasta is already preprocessed (mouth is cropped and frames extracted): $ ./predict path/to/model.h5 path/to/folder
example: $ ./predict evaluation/models/unseen-weights178.h5 evaluation/samples/bbaf2n