Portrait photos often suffer from terrible lighting and shadowing because of sub-optimal conditions in the environment, and professionals resolve this issue by adding light shaping tools, like scrims, bounce cards, as well as camera flashes. Researchers have developed a method that relies on a pair of AI-powered neural networks remove foreign shadows cast by external objects, and another to soften facial shadows cast by the features of the subject as well as adding a synthetic fill light to improve the lighting ratio.
To train the first neural network, a dataset was constructed of real-world portraits wherein synthetic foreign shadows are rendered onto
the face. Next, the second neural network used a dataset of Light Stage scans of human subjects to construct input/output pairs of input images harshly lit by a small light source, and variably softened and fill-lit output images of each face. They then proposed a way to explicitly encode facial symmetry as well as show that the dataset and training procedure enable the model to generalize to images taken in the wild. When combined, these networks produce the realistic and aesthetically pleasing enhancements in real-world images.
- Next-gen, best-in-class laptop with the versatility of a studio and tablet, so you can type, touch, draw, write, work, and play more naturally
- Faster than Surface Pro 6, with a 10th Gen Intel Core Processor – redefining what’s possible in a thin and light computer. Wireless : Wi-Fi 6:...
- More ways to connect, with both USB-C and USB-A ports for connecting to displays, docking stations and more, as well as accessory charging
