This project is ASR-IOT project: smart home from scratch
- input: audio
- output: executes the input command
due to the limitation of the hardware(CPU and memory) and compatibilities, we have to deal with challenges to optimize the process.
- memory management
- latency
- data loss
- sampling rate
- noise
ASR model:
- Google Cloud Speech-to-Text
- (optional) builds your own model using Kaldi toolkit
Hardware requirements:
- STM32F4-Discovery board
- ESP8266 (for WIFI connection)
- microphone sensor
- (optional)other IOT devices
in progress:
- speaker verification
- natural language understanding
License under the MIT License