Adobe announced its new experimental tool, Project Voco, at the latest MAX Conference in San Diego, California. It allows voice conversion, editing of speech and creating recordings easily. It even allows users to add words someone did not say into recordings.
Adobe Project Voco Latest News & Update: How Does It Work?
Project VoCo is the product formed from the collaboration of Adobe Research and Princeton University. According to Techcrunch, in Project VoCo, users just need about 20 minutes of voice samples from a speaker. The audio application analyzes the given speech and understands the makeup of the speaker's voice. It breaks the sound down into phonemes and transcribes it. Users can now utilize the created voice models to create an all-new recording.
Editing the speech is also possible through the voice models created by Project VoCo. When the fake recording is compared to the actual recording, people will have a hard time distinguishing which is true.
Project VoCo that is based on voice conversion is not based on the technology for traditional speech synthesis. Hence, there is no need for manual intervention. Users do not need to set timestamps and other things just like the traditional one. They can correct auto-generated transcript for improved synthesis and let algorithms help in the process.
According to The Verge, Adobe developer Zeyu Jin says that the Project VoCo acts like a Photoshop for audio. It has the ability to add words or sentences that someone did not say or not originally found in the audio file. It can even be used to generate words using the given voice record of the speaker.
Adobe Latest News & Update: Benefits Of Project VoCo
Project VoCo benefits users as it can add lacking words or replace mistaken words in someone's speech. This can be used to polish clips and clean up voice recordings and podcasts. Adobe also announced two new projects for editing. It includes Project Quick Layout for editing print layouts easily and Project Clover for VR editing.