This application can recognize users' emotion from photo or speech using ML models. It will recommend some songs that are fit to users' emotions. If users fail into bad emotions many times, it will give users a warning. It is good for users' and comunities' mental health.
Disclaimer:
- In speech_text.ipynb, I refer to these two sources: https://www.thepythoncode.com/article/using-speech-recognition-to-convert-speech-to-text-python https://github.com/amrrs/emotion-detection-from-text-python/blob/main/Text_Emotion_Detection_in_Python.ipynb
- In emotion_recog.ipynb, I refer to this source: https://www.kaggle.com/code/anisbouaziz/emotion-recognition. The dataset is fer2013: https://www.kaggle.com/datasets/deadskull7/fer2013
Instruction:
- We cannot upload our saved model for image classification due to size limit. Please run emotion_recog.ipynb file first and save the model xcep.h5 We provide a Google Drive link for the model: https://drive.google.com/file/d/1hNq1qsNgjeD1G9fDgGtFIhcGGKA_ChGZ/view?usp=sharing
- To run the web application, open a terminal/cmd in this folder and run "streamlit run deplo.py"
- Please check the required libraries for the web application in requirement.txt
- PPT: https://docs.google.com/presentation/d/1N9LmbChJVLTvkGsmS4XmFK9q3yDGBGfBTTWan9RJrPo/edit#slide=id.g10365d5da27_0_379