The Realization Of A Simple Speech-dialogue System Based On Google Speech-API

Speech recognition is a voice communication technology, which takes the voice as theresearch object. The machine transfers voice signals into the appropriate texts or commandsby identifying and understanding the process. Speech recognition involves many differentfields, such as physiology, psychology, linguistics, computer science and signals processing,and is even related to the person’s body language. Its ultimate goal is to achieve a naturallanguage communication between man and machine.With the continuous progress of the times, the voice recognition technology has beendeveloped into an independent subject of a comprehensive human intelligence of the variousstudies and is gradually becoming the key technology of the IT man-machine interface. In ourlives, interpersonal dialogue function is particularly important in the future and the future ofsocial electronic intelligence inevitable development trend. Voice, as a means for people andmachines interaction, gradually has been widely used, and how we can make the machine abetter understanding of the human voice, which is our current pressing problems.Easysiri is a voice control software, through which a man-machine dialogue can beachieved by inputting the speech through a microphone, detecting the speech and recognizingafter the action feedback. It uses Microsoft ’s Vocie-api as voice input digital, as well as thecurrent mainstream Google company speech-api as the speech text engine. The software hasthe advantage that it can not only respond to the speech input but also analyze the content ofspeech. Users get the results they want, and even the users want to know the result of theproblem. In the last chapter of the paper, speech display results are presented through specificexamples, and practical speech feedbacks are analyzed through different speech examples ofphonetic system software.