An image or video can be spoiled if there is a distracting object in the foreground, but a new artificial intelligence tool can help by digitally removing the unwanted obstruction. The AI can take out fences, raindrops and reflections in a window, and it might eventually be available on smartphones. The idea came to Jia-Bin Huang at Virginia Tech University when he visited a zoo and was frustrated at his inability to get a good photo of the animals. He and his colleagues developed their obstruction-removal algorithm to clean up such images. The AI鈥檚 neural network analyses several frames in a movie, or in a 鈥渕otion photo鈥� taken by some smartphones, and identifies the various objects in the image. It then uses any slight change of angle between frames to calculate the distance to each object. 鈥淲hen you move the camera slightly, you鈥檒l find the motion for each layer will be different because they鈥檙e at different depths,鈥� says Huang. The algorithm then separates out the objects into different layers and removes the foreground layer, providing an unobstructed view of the objects behind. [video_player id=鈥漈NZO7GWF鈥� access_level=鈥漞veryone鈥漖 Removing objects from images isn鈥檛 new, but this method of doing so is. Rather than labelling certain features as things to be removed or not, the neural network automatically discovers the distracting foreground objects in the process of learning. It is also less computationally draining. 鈥淲hat this paper does really well and nicely is to use the same approach to resolve multiple image-enhancement problems, like removing reflections from windows as well as a fence obstructing a view behind,鈥� says Dima Damen at the University of Bristol, UK, who calls it a 鈥渟eminal paper in the field鈥�. Huang hopes to make further improvements to the tool. While the three minutes it takes to decipher a 1296-by-864-pixel image is quicker than prior methods, he wants it to be improved and to run on smartphones, where it is more likely to be utilised. Reference:
