emotion subset neutral,anger,fear,happiness,sadness

[do-sa]

i want to use emotion recognition in these emotion subset neutral,anger,fear,happiness,sadness , how can i create this h5 and h5d5 files of this emotion subset

[pakzan]

As suggested in EmoPy, you can train your own model using python convolutional_model.py with dataset from Microsoft FER+. This should create the required files.

To train custom emotion subset, before model training, create dataset folders following the format in sample_image_directory

[do-sa]

i have macos sierra, can i use emopy in this os ? i understand that it doesnt work on windows

[pakzan]

emopy is a python toolkit and should work with different OS, I am using Windows 10 and it's working for me

[do-sa]

can we make your script executable

[pakzan]

sure

[do-sa]

i want to use emotion recognition in these emotion subset neutral,anger,fear,happiness,sadness as i mentioned before. i am having trouble in traning my own model with Emopy.i am very thankful if can you update your interapp app with these emotions

[do-sa]

can i join two emotion subsets in face2emo

[pakzan]

yes you can

[do-sa]

i am still having trouble in making the emo set sad,happy,anger,fear,neutral by your instructions in emopy just from face

[do-sa]

i change your example to just from face, so i use two different emo2face py, with combining two subsets to get my final sad,happy,fear,neutral and anger subset.but i still got errors

[do-sa]

yes you can i use py exe but i got this error while transferring main.py to main.exe auto py exe filenotfounderror: [errno 2] no such file or directory: './models/conv_emotion_map_036.json'

[do-sa]

can we put the selected area just only the camera in main interapp frame just above the progress bars ,how can we do it

[do-sa]

As suggested in EmoPy, you can train your own model using python convolutional_model.py with dataset from Microsoft FER+. This should create the required files.

To train custom emotion subset, before model training, create dataset folders following the format in sample_image_directory

do we need to change the model in fermodel.py too

[do-sa]

can we make this api as a web extension, open it in chrome or firefox

[Deren55]

i am interested in your api, how can we convert it to as a flask api to embed it to a web page