A demonstration of voice cloning technology was conducted during the Open Campus event held in August 2023. The system enabled participants to experience the generation of synthetic speech that closely approximated their own vocal characteristics.
Voice cloning is an advanced speech synthesis technique that leverages artificial intelligence to replicate the unique vocal traits of a specific individual. This technology facilitates the generation of speech that sounds as if it were spoken by the target individual, even in the absence of prior articulation of the specific content.
The demonstration was carried out in accordance with the following procedure:
1. Participants were instructed to read a predefined script aloud, and their speech was recorded via a web browser on a notebook computer.
2. The recorded audio was subsequently encoded into a bitstream format and transmitted to a remote server.
3. On the server, machine learning algorithms were employed to construct a personalized voice model based on the input data.
4. Upon completion of model training, participants were prompted to select one of several predefined text samples.
5. The selected text was synthesized and audibly rendered using the generated voice model, thereby allowing participants to evaluate the degree of similarity between their natural voice and the synthesized output.
The demonstration successfully elicited considerable interest from attendees, indicating a strong level of engagement and public curiosity regarding the capabilities and implications of personalized voice synthesis technologies.
