We, human beings, can imagine sounds by taking a glance at a photo: A scene of a beach may bring the sound of crashing waves to mind. You may hear sounds of horns and street advertising from a picture of a busy crossing. “Imaginary Soundwalk” is a sound installation focusing on such our unconscious behavior.
In this installation, an AI system trained with hours of video data keeps generating imaginary “soundscape” by selecting the best-matched sound to given scenery from a pre-collected massive environmental sound dataset. Viewers can freely walk around in the soundscape imagined by the AI on Google Street View. A slight gap before playing sounds nudges viewers to imagine by themselves.
By externalizing such synesthetic thinkings as a system, we tried to shed lights on the power of imagination we all have in common.

Concept/Machine Learning: Nao Tokui
UI Design/Programming: Shoya Dozono
Machine Learning: Yuma Kajihara
Network Programming: 2bit
Assistant Programmer: Robin Jungers
Various sound files are used under Creative Commons Licenses. See http://imaginarysoundscape.qosmo.jp/
Speaker Provided by sonihouse