Talent creates a strange camera that looks like a loudspeaker: construct an image of the surrounding environment by listening to sound

CTOnews.com, June 4 / PRNewswire-Asianet /-- Mexican creator and developer Diego Trujillo Pisanty recently designed a unique camera that does not capture light through a lens, but builds images of its surroundings by listening to sound. The camera uses a raspberry pie 3B module and an artificial neural network to convert sound into vision.

The shape of the camera is similar to that of a loudspeaker, with a huge trumpet mouth at the front to amplify the sound collection. Users only need to point the trumpet in the direction they want to shoot, press the button, and the camera will generate a picture through artificial intelligence based on the captured sound. The camera is also equipped with a small screen that shows the processing progress and preview of the picture.

Pisanty says he developed a customized artificial neural network (Artificial Neural Network, or ANN) for the project and trained it with videos he shot around Mexico City. He matched each frame of video with the corresponding one-second sound to establish a connection between the sound and the video. Therefore, the images generated by this camera are based on images from inside Mexico City.

Pisanty says he wants to explore the relationship between sound and vision, and how artificial intelligence understands and creates the world. He believes that the project is an art experiment as well as a technical challenge. He hopes the camera will inspire people to think about the connections and differences between sound and image.

Pisanty shared a video of the project on YouTube and provided more details on his website, which can be seen by interested CTOnews.com partners.