Use the Google speech-to-text API to generate presentation slides as you talk! A collaboration with Marc Mailhot. Marc wrote the slide renderers and Bing Search with PyGame, and I wrote the Google Voice API interfacing code and the code to process the speech, decide on slide content and tie together the modules. Made for TerribleHack IV, a hackathon for intentionally funny and stupid projects.
Includes two themes: 90s and corporate, as well as three slide types: heading, bullet points, and picture.
It can use Bing image search to find images based on what you say.
Note: This was a fun 5-hour project for a fun hackathon, it stops working and crashes often due to multithreading issues we didn't bother to debug and isn't actually useful.
See here:
https://www.youtube.com/watch?v=chgOZumnXQo
The voice recognition quality in the video is rather bad because it was a noisy room and I was using a crappy microphone.
- Set up a virtualenv if you want and know what that is.
- Install portaudio for pyAudio
- Install the dependencies of pygame, mostly SDL and related packages, see the pygame website for this.
pip install -r requirements.txt- Set yourself up with a Google Cloud Speech API account and a Bing Image Search API key
export GOOGLE_APPLICATION_CREDENTIALS=/path/to/your/credentials-key.jsonexport BING_API_KEY=yourkeygoesherepython main.py- Start saying things!
