Abstract: This specification describes a simulation system that performs simulations of physical environments using a graph neural network. At each of one or more time steps in a sequence of time steps, the system can process a representation of a current state of the physical environment at the current time step using the graph neural network to generate a prediction of a next state of the physical environment at the next time step. Some implementations of the system are adapted for hardware GLOBAL acceleration. As well as performing simulations, the system can be used to predict physical quantities based on measured real-world data. Implementations of the system are differentiable and can also be used for design optimization, and for optimal control tasks.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an action selection policy neural network, wherein the action selection policy neural network is configured to process an observation characterizing a state of an environment to generate an action selection policy output, wherein the action selection policy output is used to select an action to be performed by an agent interacting with an environment. In one aspect, a method comprises: obtaining an observation characterizing a state of the environment subsequent to the agent performing a selected action; generating a latent representation of the observation; processing the latent representation of the observation using a discriminator neural network to generate an imitation score; determining a reward from the imitation score; and adjusting the current values of the action selection policy neural network parameters based on the reward using a reinforcement learning training technique.

Abstract: A reinforcement learning system and method that selects actions to be performed by an agent interacting with an environment. The system uses a combination of reinforcement learning and a look ahead search: Reinforcement learning Q-values are used to guide the look ahead search and the search is used in turn to improve the Q-values. The system learns from a combination of real experience and simulated, model-based experience.

Abstract: An improved system to provide a clothing model in augmented reality and/or virtual reality is disclosed. In one embodiment, the system provides a fully simulated garment, on the user's device, in an augmented reality or virtual reality display. In one embodiment, the process uses a parameterized smart garment model which enables real-time interaction with the garment on a device such as a mobile phone or tablet, with limited memory and processing power. In one embodiment, the system permits nearly instantaneous alteration of garment sizes. Because the garment is modeled to accurately reflect physical details, the lay of the garment changes as the size of the garment or body changes.

Abstract: The present invention relates to building a computer model of a garment based on a physical sample garment, and to the process of using the computer model of a garment to determine the garment's appearance on a human body.

Abstract: The present invention relates to building a computer model of a garment based on a physical sample garment, and to the process of using the computer model of a garment to determine the garment's appearance on a human body.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an action selection policy neural network, wherein the action selection policy neural network is configured to process an observation characterizing a state of an environment to generate an action selection policy output, wherein the action selection policy output is used to select an action to be performed by an agent interacting with an environment. In one aspect, a method comprises: obtaining an observation characterizing a state of the environment subsequent to the agent performing a selected action; generating a latent representation of the observation; processing the latent representation of the observation using a discriminator neural network to generate an imitation score; determining a reward from the imitation score; and adjusting the current values of the action selection policy neural network parameters based on the reward using a reinforcement learning training technique.

Abstract: An improved system to provide a clothing model in augmented reality and/or virtual reality is disclosed. In one embodiment, the system provides a fully simulated garment, on the user's device, in an augmented reality or virtual reality display. In one embodiment, the process uses a parameterized smart garment model which enables real-time interaction with the garment on a device such as a mobile phone or tablet, with limited memory and processing power. In one embodiment, the system permits nearly instantaneous alteration of garment sizes. Because the garment is modeled to accurately reflect physical details, the lay of the garment changes as the size of the garment or body changes.

Abstract: A method and apparatus to provide a depiction of a garment model of a body shape is described. The garment model takes into account the visual and mechanical characteristics of the fabric, as well as the drape of the fabric on a body shape.