I like coding, playing chess, and long walks on the beach. Don't worry, I'm definitely not making AI to take over the world. 🤖
✅ Masters in AI at Deakin University
✅ Worked with speech, text, and image data
✅ Data Visualization and Analysis 📈
✅ Looking for jobs 👀
✅ Full working rights in Australia 🇦🇺
❌ Can carry you in video games
Here's a random video of me coding the robot Pepper doing air guitar during my time as a Research Assistant at Deakin.
- The first problem was that Pepper picked up fan noises and any other background noises which the speech-to-text libraries did not appreciate. Spectral Gating Noise Reduction was implemented via this library in order to fix this problem.
- For speech-to-text, we tested with CMU Sphinx, Google Speech Recognition, Wit.ai and IBM Speech to Text, and finally decided to stick with Google Speech Recognition.
- All of this was to be sent back and forth between python versions using a flask server as Pepper works on an older python version.
- This text was inputted into GPT4All and we used retrival-augmented generation using Deakin's courses so that it could correlate and find courses which the user would like.