Deep-learning Technique Reveals Objects in Pitch-Dark Condition

Massachusetts Institute of Technology (MIT) Engineers successfully trained a deep neural network to detect tiny objects in the dark that are normally invisible to human eyes.

Deep learning tries to emulate the behavior of the human brain to process input data sets. It consists of different layers of mathematical models and each accepting input and predicting the output at every stage. The inputs are generally fed in the form of images, text, and sound. Its performance and accuracy often surpass human-level.

The study is published in Physical Review Letters where the researchers reconstructed transparent objects such as Small imperfections in a wine glass or tiny creases in a contact lens, in almost pitch-black conditions. Barbastathis’ co-authors on the paper are lead author Alexandre Goy, Kwabena Arthur, and Shuai Li.

Co-authors on the paper, Alexandre Goy said, “When we look with the naked eye, we don’t see much — they each look like a transparent piece of glass. But there are actually very fine and shallow structures that still have an effect on light.”

From an original transparent etching (far right), engineers produced a photograph in the dark (top left), then attempted to reconstruct the object using first a physics-based algorithm (top right), then a trained neural network (bottom left), before combining both the neural network with the physics-based algorithm to produce the clearest, most accurate reproduction (bottom right) of the original object. Courtesy of the MIT researchers

The 10,000 transparent images of glass-like etchings, based on extremely grainy patterns, were used to train the neural network. The images were taken in very low lighting conditions, with about one photon per pixel. This is far less light than a camera would register in a dark sealed room.

As a use case, transparent features of biological tissues and cells can be illuminated with the help of this technique.

Professor of mechanical engineering at MIT, George Barbastathis, said, “In the lab, if you blast biological cells with light, you burn them, and there is nothing left to the image. When it comes to X-ray imaging, if you expose a patient to X-rays, you increase the danger they may get cancer. What we’re doing here is, you can get the same image quality, but with a lower exposure to the patient. And in biology, you can reduce the damage to biological specimens when you want to sample them.”