Speech Recognition/In-Car Systems
Learning Tasks
[edit | edit source].jpg)
In cars typically a manual control input is used. Define control elements, for which by means of finger or foot control manual control is required from your point of view? What are the requirements and constraints to replace the steering-wheel by voice control ("turn left", "turn right", "go to Baker Street 6", ...). If you enable the speech recognition system noise in the car can reduce the quality of recognition. When is a speech recognition a security risk and when does speech recognition improve the security?
- (Voice Control Music) Start with simple voice commands for phone calls, select radio stations or play music from a compatible smartphone, MP3 player or music-loaded flash drive.
- (Comparision Switch on Lights and Switch on Sound System) To "switch on lights" with voice recognition might be similar to switching the radio by voice control. What are the differences?
- (Speech Recognition - Voice Changes) If voice changes due a cough the voice recognition might fail. How should a voice recognition deal with that challenge?
- (Mathematical comparision Slow Down Speed. and Reduce Volumn Sound System) From the mathematical difference there is not a big difference between changing the volume von 25% to 70% in comparision to alteration of speed from 25mph to 70mph. Is it possible to used one speech recognition system for both applications? How would you design the commands for the speech recognition system to deal with both cases?
- (Javascript and Speechrecognition) Explore PocketSphinx Javascript as OpenSource application. What are the benefits and challenges of Speech recognition with webbased technology (HTML, CSS, Javascript) and could these technologies be used with in-car systems. Would you prefer remote speech recognition on servers or local on-board speech recognition systems without internet connectivity?