Robot arm control by using voice commands

Recognition systems, which perform remote control of devices, have gained significant importance day by day. Speech, face, and fingerprint recognition systems seem to be the most frequently used recognition systems. Speech recognition systems can be used in real time in security systems, device control systems and dictation systems. In this study, the robot arms controlling by recognizing the real-time speech commands have been done. The Fisher Linear Discriminant Analysis (FLDA) and Discriminative Common Vector Approach (DCVA) of effective subspace classifiers have been used to recognize speech commands. One of these classifiers, DCVA was used for the first time in voice recognition. For the training set, a total of 24 speech sentences have been created for four different objects which have got six different colors. The training set has been created as speaker-dependent. During the test phase, by recognizing the speech signals in real-time, the robot arm has been directed to the object whose coordinates have been previously determined. In order to perform this direction process, firstly speech commands are recognized in real time by the computer software interface and the relevant data are transmitted to the robot control card using the RS232 serial communication protocol according to the recognized command. Then, with the help of the microcontroller containing the information of the location of each object on the control card, the robot's servo motors are directed towards the object position. As a result of the study, the average speech recognition rate for DCVA with language model was 98.3% and without language model was 90,73%. For the FLDA, the average speech recognition rate without language model was 89.48% and with language model was 97.1%. For ANN, the average speech recognition rate with language model was 91,1% and without language model 81,9%. Keywords: Speech recognition, Robot arm controlling, FLDA, DCVA, ANN