Joy - Voice & Vision Assistant

Note: This serves as the product page for Joy where I keep track of all the feature and enhancement updates

Joy is a virtual assistant which comes with both voice and vision capabilities. Originally started with the idea of aiding those with vision challenges, Joy is now generalized and powered by Rasa for its conversation handling and several Machine Learning models developed using TensorFlow and Keras for providing vision related features.

Virtual assistants are all around us(already!) and their number is expected to grow in the coming years. While we all are accustomed to using touch for interacting with devices, having voice as the interface is going to add a lot of value to the existing systems, especially in aiding those with vision challenges through voice and speech. We already have products from the big tech companies who are betting big on voice assistants like Google Assistant, Alexa, Cortana etc. While these virtual assistants’ presence has already reached the common households, it is necessary to have more participants in this area. While it is fun, easy and handy to use virtual assistants in our phones and at home, building one ourselves is both a challenging and time taking task. As I started to work on building one such virtual assistant from scratch, I got the chance to understand the underlying technology and process.