Just as some people are amazingly good at focusing their attention on a particular person in a noisy environment, mentally “muting” all other voices and sounds, it may soon be possible to pick out individual voices in a crowd by suppressing all other sounds, thanks to a new Artificial Intelligence (AI) system developed by Google’s AI researchers.
Known as the cocktail party effect, the capability to mentally "mute" all other voices and sounds comes natural to us humans. However, automatic speech separation — separating an audio signal into its individual speech sources — while a well-studied problem, remains a significant challenge for computers, said Inbar Mosseri and Oran Lang, software engineers at Google Research, wrote in a blog post this week.
In a new paper titled “Looking to Listen at the Cocktail Party”, the AI engineers presented a deep learning audio-visual model for isolating a single speech signal from a mixture of sounds such as other voices and background noise.
"In this work, we are able to computationally produce videos in which speech of specific people is enhanced while all other sounds are suppressed," Mosseri and Lang said. The method works on ordinary videos with a single audio track, and all that is required from the user is to select the face of the person in the video they want to hear, or to have such a person be selected algorithmically based on context.
The researchers believe this capability can have a wide range of applications, from speech enhancement and recognition in videos, through video conferencing, to improved hearing aids, especially in situations where there are multiple people speaking.
"A unique aspect of our technique is in combining both the auditory and visual signals of an input video to separate the speech," the AI engineers said. "Intuitively, movements of a person's mouth, for example, should correlate with the sounds produced as that person is speaking, which in turn can help identify which parts of the audio correspond to that person.”
The visual signal not only improves the speech separation quality significantly in cases of mixed speech, but, importantly, it also associates the separated, clean speech tracks with the visible speakers in the video, the researchers noted.