NowHere Media is an award-winning creative studio that produces virtual and augmented reality experiences. They develop impactful stories which inspire and engage global audiences. Their recent production Home After War was toured in prestigious film festivals all over the world and won the Social Impact Media Award 2019.
NowHere Media is currently working on a VR project. For this production, they needed a speech-to-text AI system. Users can use the speech recognition system to navigate the VR experience with vocal instructions.
We built a prototype speech recognition system using deep learning techniques. We trained the AI using audio recordings of words . Once we developed the system, we deployed it in the cloud using Amazon servers. The speech recognition system can be queried from the VR experience through an API endpoint. The VR experience sends an audio file to be analysed on the cloud. The speech recognition system processes the audio and returns the predicted word to the VR experience.