Applications Old 2023 - FEMTOSENSE

WHAT WE’VE ENABLED

Real-time speech enhancement <1 mW

Intelligent transparency mode and hearing enhancement for earbuds, headsets, and hearing aids

Always-listening keyword detection <100 µW

Hands-free remote control that lasts over 2 years on batteries

Always-on multiclass sound event detection

Intelligent security system detecting glass breaking, baby crying, gunshots, and alarms at the sensor nodes.

WHAT WE CAN ENABLE

SPU-001

Speech/Audio
- Neural Beamforming
- Scene ID
- Sound Classification
- Anomaly Detection
- Source Separation
- Custom Applications

SPU-002

COMING SOON

Speech/Audio
- Neural Beamforming
- Scene ID
- Sound Classification
- Anomaly Detection
- Source Separation
- Custom Applications
Vision/Video
- Image/Video Super Resolution
- Object Recognition
- Object Tracking
- Image Segmentation
- Face Detection/ID
- Image/Video Generation
- Custom Applications
NLP
- Automatic Speech Recognition
- Language Detection
- Language Translation
- Text-to-speech
- Sentiment Analysis
- Dialog Generation/Chatbots
- Custom Applications

COMING SOON

Speech to Text

Femtosense’s dual sparsity design consumes much less energy than existing approaches by using only a fraction of active neurons.

AI Noise Cancellation

Femtosense showcased an ultra-low-latency speech enhancement demo. Contact us to experience our latest version in-person and believe what you hear!