We propose a new family of policy gradient methods for reinforcement learning, which alternate between sampling data through interaction with the environment, and optimizing a "surrogate" objective function using stochastic gradient ascent.
Recent advances in machine learning are consistently enabled by increasing amounts of computation.
Motivated by this, a number of artificial intelligence (AI) systems based on deep learning have been proposed and results have been shown to be quite promising in terms of accuracy in detecting patients infected with COVID-19 using chest radiography images.
This large scale study focuses on quantifying what X-rays diagnostic prediction tasks generalize well across multiple different datasets.
We present a novel method for monocular hand shape and pose estimation at unprecedented runtime performance of 100fps and at state-of-the-art accuracy.
Recently there has been an interest in the potential of learning generative models from a single image, as opposed to from a large dataset.
We introduce SinGAN, an unconditional generative model that can be learned from a single natural image.