Why is that we still need others to take pictures for us if we want to see ourselves outside of the selfie context? Why must we resort to using camera timers? If we were able to make a camera app completely controlled by voice recognition, these problems would be solved!
What it does
It is a simple camera app that uses voice recognition to take photos, adjust several parameters such as zoom, exposure, etc.
How we built it
We used unity to build and android app that uses the CMU Sphinx voice recognition API.
Challenges we ran into
Finding reliable speech recognition API. CMU Sphinx API is not very sophisticated and interprets a lot of noise as words. We tried to implement filters, but we were unable to find efficient and reliable algorithms to modify the images in real time.
Accomplishments that we're proud of
Creating an android app that is capable of taking pictures. Implementing speech recognition in a mobile app. Using unity to make a simple UI that demonstrates the app's recognition of the user's instructions.
What we learned
Advanced filters are really complex, and requires advanced machine learning. How to use Unity to develop an android app (two out of our 3 members had very little experience with Unity).
What's next for Cammand
Fine tuning the voice recognition Adding more features like adjusting the zoom, exposure, filters.