Lab view based- voice activation system for switching home appliances utilizing audio fingerprinting algorithm

• Main Authors: Aco, Joseph Emil F., Go, Margaret D., Nulud, Reginald Noel P., Pedron, Jonathan Luigi R.
• Format: text
• Language: English
• Published: Animo Repository 2012
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/14800
• Institution: De La Salle University

This thesis aims to create a voice-activated home automation system that will control four appliances (television, lights, roll-up/roll-down blinds and remote-controlled electric fan). Relay switches were also integrated in the system in controlling certain appliances opto-isolators served as circuit protection. A software called LabVIEW was also used especially the LabView FPGA Module, which allowed the group to use the output ports of the Spartan 3E FPGA Board. Audio Fingerprinting method was used for the speech recognition process. This was done by converting the analog voice signal into digital form, then it will be filtered out by using eight fourth-order Butterworth bandpass filters, which serve as filterbanks. These eight filters pass frequencies with an increment of 200 Hz range from 200 Hz up to 1800 Hz. The system was also tested and implemented in an unconstrained environment. Results showed that the system can recognize words even if the microphone and the speaker are distant from each other. This was attained by utilizing an array microphone to obtain the voice input from the user. The voice recognition proved to be speaker-independent. However, it is more efficient when voice recognition is speaker-dependent.