Abstract: A game platform manages, on behalf of a game program, platform nodes used to execute game operations, such as physics operations, artificial intelligence operations, and player messaging operations. Via the game platform, the game program can implement relatively sophisticated and complex game operations. The game platform can implement each game operation at a corresponding platform node, wherein the platform node employs associated hardware to execute the game operation. The game platform can distribute the platform nodes at different sets of hardware resources, such as different servers, different hardware accelerators, and the like, and thereby implement more sophisticated game operations than can be implemented by the game program directly. The game platform communicates data between the game program and the platform nodes using a zero-copy transfer technique, thereby transferring data with relatively little latency.

Abstract: A memory stores information representing a set of canonical utterances. A processor receives information representing an utterance from a first user of an application and selects a canonical utterance from the set of canonical utterances based on semantic comparisons of the utterance from the first user and the set of canonical utterances. The semantic comparisons include semantic retrieval and semantic similarity operations that can be performed by a semantic natural language processing machine learning model. The processor presents the canonical utterance to a second user of the application instead of presenting the utterance from the first user. In some cases, the processor replaces the utterances from the user in a text stream or a voice chat with the canonical utterances in the set of canonical utterances.

Abstract: A virtual camera captures first images of a three-dimensional (3D) digital representation of a visual asset from different perspectives and under different lighting conditions. The first images are training images that are stored in a memory. One or more processors implement a generative adversarial network (GAN) that includes a generator and a discriminator, which are implemented as different neural networks. The generator generates second images that represent variations of the visual asset concurrently with the discriminator attempting to distinguish between the first and second images. The one or more processors update a first model in the discriminator and/or a second model in the generator based on whether the discriminator successfully distinguished between the first and second images. Once trained, the generator generates images of the visual asset based on the first model, e.g., based on a label or an outline of the visual asset.

Abstract: Systems and methods are provided for receiving at least one image and a reference image, and performing a plurality of downscaling operations having separable convolutions on the received at least one image. A plurality of residual blocks may be formed, with each residual block containing two separable convolutions of the kernel and two instance normalizations. A plurality of upscaling operations may be performed on the plurality of residual blocks, and a stylized image may be displayed based on at least the performed plurality of upscaling operations and the reference image.