Abstract: A method for delivering interactive video to a user is disclosed herein. A computing system identifies video contents corresponding to a sporting event. The video contents include a plurality of video frames. The computing system annotates each video frame of the plurality of video frames to uniquely identify the video frame and contents contained therein. The computing system receives, from a plurality of prediction models, a plurality of data inputs related to agents and actions captured in each video frame of the plurality of video frames. The computing system generates a plurality of data frames based on the plurality of data inputs. The computing system associates each data frame with a respective video frame using the annotations. The computing system causes a user device to present the interactive video to the user by instructing the user device to merge the plurality of data frames with the plurality of video frames.

Abstract: Hardware accelerator designs for neural networks are improved with various approaches to reduce circuit area, improve power consumption, and reduce starvation. Convolutional layers of a neural network may multiply a set of weights with a set of inputs. One example defers two's complement arithmetic from the parallelized multiplication circuits and completes the two's complement arithmetic when the results are accumulated. In another example, a multiplication circuit initially multiplies an input by an initial value of the maximum (or minimum) multiplication range before applying the magnitude of a multiplication encoded relative to the multiplication range. In another example, after dimensional reduction earlier in the network hardware, circuitry for a convolutional layer uses a reduced number of convolutional block circuits that are reused across a plurality of clock cycles to apply different subsets of weight channels.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed for hardware-aware machine learning model training. An example apparatus includes a configuration determiner to determine a hardware configuration of a target hardware platform on which the machine learning model is to be executed, a layer generator to assign sparsity configurations to layers of the machine learning model based on the hardware configuration, and a deployment controller to deploy the machine learning model to the target hardware platform in response to outputs of the machine learning model satisfying respective thresholds, the outputs including a quantity of clock cycles to execute the machine learning model with the layers having the assigned sparsity configurations.

Abstract: A bidirectional connection between a multimedia stream distribution system and a user is used to communicate operational data to and from the user. The bidirectional connection is separate from a multimedia connection that transfers a multimedia stream from the multimedia stream distribution system to the user. Providing a bidirectional connection separate from the multimedia connection is advantageous, among other reasons, because the operational data can be collected effectively at the multimedia stream distribution system independent of any limitations in the multimedia connection. To organize users and reduce processing load, the multimedia stream distribution system assigns users to user groups based on shared characteristics. Groups may be used to propagate information to the users as well as analyze information collected from the users in an efficient manner.

Abstract: A bidirectional connection between a multimedia stream distribution system and a user is used to communicate operational data to and from the user. The bidirectional connection is separate from a multimedia connection that transfers a multimedia stream from the multimedia stream distribution system to the user. Providing a bidirectional connection separate from the multimedia connection is advantageous, among other reasons, because the operational data can be collected effectively at the multimedia stream distribution system independent of any limitations in the multimedia connection. To organize users and reduce processing load, the multimedia stream distribution system assigns users to user groups based on shared characteristics. Groups may be used to propagate information to the users as well as analyze information collected from the users in an efficient manner.

Abstract: A bidirectional connection between a multimedia stream distribution system and a user is used to communicate operational data to and from the user. The bidirectional connection is separate from a multimedia connection that transfers a multimedia stream from the multimedia stream distribution system to the user. Providing a bidirectional connection separate from the multimedia connection is advantageous, among other reasons, because the operational data can be collected effectively at the multimedia stream distribution system independent of any limitations in the multimedia connection. To organize users and reduce processing load, the multimedia stream distribution system assigns users to user groups based on shared characteristics. Groups may be used to propagate information to the users as well as analyze information collected from the users in an efficient manner.

Abstract: A bidirectional connection between a multimedia stream distribution system and a user is used to communicate operational data to and from the user. The bidirectional connection is separate from a multimedia connection that transfers a multimedia stream from the multimedia stream distribution system to the user. Providing a bidirectional connection separate from the multimedia connection is advantageous, among other reasons, because the operational data can be collected effectively at the multimedia stream distribution system independent of any limitations in the multimedia connection. To organize users and reduce processing load, the multimedia stream distribution system assigns users to user groups based on shared characteristics. Groups may be used to propagate information to the users as well as analyze information collected from the users in an efficient manner.