Inspiration
With concerns for mental health at an all-time high, and therapy still inaccessible to most, be it due to societal or financial concerns, we have realized the need for accessible therapy that doesn’t leave you desperate to buy a rope and too destitute to do it.
What it does
With the help of machine learning, We intend to create a smart therapist to assist users on a daily basis by analyzing the emotions and the inputs given to the machine.
How we built it
We are using TensorFlow, nltk, to train and make an efficient therapist chatbot, and using the different emotions provided by the users, which we will be using haarcascade files and OpenCV library to detect and perform image processing/ video processing.
Challenges we ran into
It found smiles that didn't exist. The algorithm used is not 100% accurate, which utilizes different models such as mobilenet, imagenet, vgg16, xception, etc. that are relatively complex, are time-consuming and require a strong GPU. The amount of data required for this project to work is very high. The research availability takes a lot of time.
Accomplishments that we're proud of
It detects faces with appreciable accuracy. There are a lot of new things that we discovered while doing this project.
What we learned
Various modules, and usage of different layer architecture, convolutional neural networks.
What's next for Faux therapist
We intend to make it a mobile therapist, that is readily available, and helps you get through the lemons life gives you.
Log in or sign up for Devpost to join the conversation.