A team of engineers from MIT just developed a new machine learning technique that reveals “invisible” objects in complete darkness.
According to the MITÂ engineers, their new machine learning technique enables them to see even the tiniest imperfections in transparent objects.
In a paper published by the MIT engineers in the journal Physical Review Letters, the team reported how they used a deep neural network to reconstruct transparent objects from images of the same objects taken in pitch-black conditions.
The MIT researchers reportedly trained the neural network to associate specific inputs with corresponding outputs. In the team’s experiment, the inputs were the dark, grainy images of the transparent objects. They then used outputs of the objects themselves.
The New Machine Learning Technique
The MIT researchers trained the computer to recognize over 10,000 transparent glass-like etchings based on images with extremely grainy patterns. To test their new machine learning technique, the researchers took a picture in very low lighting conditions with about one photon per pixel.
After this, the team showed the grainy image not included in the training data to their computer. Surprisingly, the neural network could reconstruct the transparent object. This demonstrated that the technology can detect transparent features like cells and biological tissues even in near darkness.
“We have shown that deep learning can reveal invisible objects in the dark. This result is of practical importance for medical imaging to lower the exposure of the patient to harmful radiation, and for astronomical imaging,” Alexandre Goy, lead author of the study, said.
The Intelligence Advanced Research Projects Activity and Singapore’s National Research Foundation supported the team’s research.
Comments (0)
Least Recent