Talking to Objects 2024 - Ongoing, Dataset, Video with Lambda Vue
Computer vision models can recovers speech from the vibrations off objects. Using Lambda Vue, a software that amplifies small motions in video, and a microscope, the project captures microscopic voice markers from material vibrations. In this case study, tinfoyl (phonetic writing) is spoken to tinfoil, generating a visual dataset that matches each graphemes (letters) and phonemes (sounds).
With Caro Trigo, Mavis Yue Cao, Christie Wu
Video that visualizing the vibrations of speech. Saying tinfoil over tinfoil. Before and after Lamda Vue software processing Tinfoil grid to define spoken phenomes and visual letters. Separating each frame and phenomeTYPE SPECIMEN - Tinfoil‘T’ framesIdentifying ‘T’ sound markers and characteristics in the materialSelecting high contrast markers for each phenomeIdentifying sound markers and characteristics in the materialSound markersMicroscope setupMicroscope setup in Immersion LabVideo that visualizing the vibrations of speech on a strawberry.