Speech recognition in the car is becoming must-have not just in luxury cars. Noises like kids talking in the back and transient noises such as sirens pass thru directional microphones and noise cancelling systems. Hi Auto makes recognition systems work under any noise conditions.
Speech recognition platforms today suffer from poor recognition rates in a noisy environment such as kids in the back seat of cars, radio, ambulance passing by, or heavy rain. Video conferencing becomes ineffective when one of the participants is in a noisy environment like a coffee shop. Refer https://hi.auto/.
Hi Auto developed a software only speech enhancement and speaker separation system that eliminates the most challenging noises and focus only on the speaker.
Hi Auto’s, single microphone approach versus microphone array will save OEMs on BOM.
Hi Auto Audio-Visual speech enhancement can eliminate the most challenging noises. Certain common noises go through directional microphones, making speech recognition systems less reliable and hurt the quality of calls. Common problems:
1. Potential Word Error in excess of 70%.
2. Also, short strong transient noises go through directional microphones and throw off speech recognition systems.
3. Kids talking in the back goes through directional microphones and throws off speech recognition systems.
In current and future Semi-Autonomous Vehicles & People Movers Voice quality is a key enabler for 3 areas:
– Sharing space with others requires focus on the right speaker.
– Reliability is key for safety.
– Provide the user a better embedded experience than the projected phone experience.
The needs are:
1. Ride sharing & family cars need to provide personalized experience for each passenger.
2. L2+/3: Pass, change lane, Stop, Park – Car control & Driver Alertness.
3. Improved Infotainment Experience.
Hi Auto Audio-Visual speech enhancement eliminates the most challenging noises. Speech enhancement that can reduce Word Error to < 8%.
1. Hi Auto is a deep learning software only system that leverages a speaker facing camera and a single microphone for more accurate speech recognition and clear speech enhancement.
2. The system works on device and eliminates the most challenging noise sources as multiple speakers, radio, sirens, and heavy rain.
3. Provides “sound bubble” effect removing all noise.
4. Supports both voice recognition and phone calls purposes.
5. Language agnostic.
6. Software only solution that works completely embedded
– Uses existing microphone and driver facing camera
– No video is leaving the car (privacy)
– Efficient in memory and CPU
– Works with simple web or IR camera and can be adjusted to different positions.
For more information contact Roy Baharav, CEO & Co-founder, roy@hi.auto.
