Abstract: According to one implementation, a pixel error detection system includes a hardware processor and a system memory storing a software code. The hardware processor is configured to execute the software code to receive an input image, to mask, using an inpainting neural network (NN), one or more patch(es) of the input image, and to inpaint, using the inpainting NN, the masked patch(es) based an input image pixels neighboring each of the masked patch(es). The hardware processor is configured to further execute the software code to generate, using the inpainting NN, a residual image based on differences between the inpainted masked patch(es) and the patch(es) in the input image and to identify one or more anomalous pixel(s) in the input image using the residual image.

Abstract: According to one implementation, a system for performing automated salience assessment of pixel anomalies includes a computing platform having a hardware processor and a system memory storing a software code. The hardware processor is configured to execute the software code to analyze an image for a presence of a pixel anomaly in the image, obtain a salience criteria for the image when the analysis of the image detects the presence of the pixel anomaly, and classify the pixel anomaly as one of a salient anomaly or an innocuous anomaly based on the salience criteria for the image. The hardware processor is further configured to execute the software code to disregard the pixel anomaly when the pixel anomaly is classified as an innocuous anomaly, and to flag the pixel anomaly when the pixel anomaly is classified as a salient anomaly.

Abstract: According to one implementation, a pixel error detection system includes a hardware processor and a system memory storing a software code. The hardware processor is configured to execute the software code to receive an input image, to mask, using an inpainting neural network (NN), one or more patch(es) of the input image, and to inpaint, using the inpainting NN, the masked patch(es) based an input image pixels neighboring each of the masked patch(es). The hardware processor is configured to further execute the software code to generate, using the inpainting NN, a residual image based on differences between the inpainted masked patch(es) and the patch(es) in the input image and to identify one or more anomalous pixel(s) in the input image using the residual image.

Abstract: In various embodiments, a storyboarding application generates a storyboard for a media title. In operation, the storyboarding application determines a categorization for a first portion of the media item. The storyboarding application then determines a first media item based on at least one of the categorization or a caption associated with the first portion of the media title. Subsequently, the storyboarding application modifies the first media item based on at least one of the categorization or a character associated with the caption to generate a second media item. The storyboarding application then generates a sequence of media items for the storyboard that includes the second media item. Advantageously, because the storyboarding application can automatically generate media items for storyboards based on categorizations and/or captions, the storyboarding application can reduce both the manual effort and time required to generate storyboards relative to prior art techniques.

Abstract: A graphical user interface (“GUI”) personalization application automatically generates a GUI for a software application. In operation, the GUI personalization application tracks interactions between a user and a first GUI that is associated with the software application and a GUI framework. The GUI personalization application then computes a behavioral simplification for the first GUI based on the interactions and the GUI framework. Subsequently, the GUI personalization application modifies the first GUI based on the behavioral simplification to generate a second GUI for the user that is associated with the software application and the GUI framework. Advantageously, the GUI personalization application can more efficiently and reliably improve the quality of the overall user experience when interacting with the software application across a wider range of users relative to prior art approaches.

Abstract: In various embodiments, an interactive streaming application plays back a media title via a client device. In operation, the interactive streaming application causes the client device to playback a first chunk of the media title. While the client device plays back the first chunk, the interactive streaming application determines a movement of an internet of things (“IoT”) device that is controlled by the user. The interactive streaming application performs reinforcement-learning operation(s) based on the first chunk and the movement to determine a second chunk of the media title to playback. The interactive streaming application then causes the client device to playback the second chunk of the media title. Advantageously, the interactive streaming application can automatically personalize the playback of the media title for the user based, at least in part, on movements of the IoT device.

Abstract: A recommendation engine generates contextually relevant media programs tailored to the specific context of the user. The recommendation engine interfaces with one or more Internet-of-Things (IoT) appliances to determine one or more food items stored within the IoT appliance(s) and/or one or more cooking capabilities associated with the IoT appliance(s). The recommendation engine also gathers profile data that reflects a set of preferences associated with the user. Based on the gathered data, the recommendation engine queries a database of cooking program data associated with various cooking programs. The recommendation engine obtains cooking program data that is contextually relevant to the user and then generates a contextual cooking program that describes how to perform a given recipe. During playback of the contextual cooking program, the recommendation engine can modify the contextual cooking program based on the progress of the user in following the recipe.

Abstract: A data processing system includes a computing platform having a hardware processor and a memory storing a data compression software code. The hardware processor executes the data compression software code to receive a series of compression input data and encode a first compression input data of the series to a latent space representation of the first compression input data. The data compression software code further decodes the latent space representation to produce an input space representation of the first compression input data corresponding to the latent space representation, and generates f refined latent values for re-encoding the first compression input data based on a comparison of the first compression input data with its input space representation. The data compression software code then re-encodes the first compression input data using the refined latent values to produce a first compressed data corresponding to the first compression input data.

Abstract: Systems and methods are disclosed for generating a latent space residual. A computer-implemented method may use a computer system that includes non-transient electronic storage, a graphical user interface, and one or more physical computer processors. The computer-implemented method may include: obtaining a target frame, obtaining a reconstructed frame, encoding the target frame into a latent space to generate a latent space target frame, encoding the reconstructed frame into the latent space to generate a latent space reconstructed frame, and generating a latent space residual based on the latent space target frame and the latent space reconstructed frame.

Abstract: According to one implementation, an image compression system includes a computing platform having a hardware processor and a system memory storing a software code. The hardware processor executes the software code to receive an input image, transform the input image to a latent space representation of the input image, and quantize the latent space representation of the input image to produce multiple quantized latents. The hardware processor further executes the software code to encode the quantized latents using a probability density function of the latent space representation of the input image, to generate a bitstream, and convert the bitstream into an output image corresponding to the input image. The probability density function of the latent space representation of the input image is obtained based on a normalizing flow mapping of one of the input image or the latent space representation of the input image.

Abstract: Various embodiments disclosed herein provide techniques for performing incremental natural language understanding on a natural language understanding (NLU) system. The NLU system acquires a first audio speech segment associated with a user utterance. The NLU system converts the first audio speech segment into a first text segment. The NLU system determines a first intent based on a text string associated with the first text segment, wherein the text string represents a portion of the user utterance. The NLU system generates a first response based on the first intent prior to when the user utterance completes.

Abstract: Systems and methods are disclosed for reconstructing a frame. A computer-implemented method may use a computer system that includes non-transient electronic storage, a graphical user interface, and one or more physical computer processors. The computer-implemented method may include: obtaining one or more reference frames from non-transient electronic storage, generating one or more displacement maps based on the one or more reference frames and a target frame with the physical computer processor, generating one or more warped frames based on the one or more reference frames and the one or more displacement maps with the physical computer processor, obtaining a conditioned reconstruction model from the non-transient electronic storage, and generating one or more blending coefficients and one or more reconstructed displacement maps by applying the one or more displacement maps, the one or more warped frames, and a target frame to the conditioned reconstruction model with the physical computer processor.

Abstract: Systems and methods are disclosed for reconstructing a frame. A computer-implemented method may use a computer system that includes non-transient electronic storage, a graphical user interface, and one or more physical computer processors. The computer-implemented method may include: obtaining one or more reference frames from non-transient electronic storage, generating one or more displacement maps based on the one or more reference frames and a target frame with the physical computer processor, generating one or more warped frames based on the one or more reference frames and the one or more displacement maps with the physical computer processor, obtaining a conditioned reconstruction model from the non-transient electronic storage, and generating one or more blending coefficients and one or more reconstructed displacement maps by applying the one or more displacement maps, the one or more warped frames, and a target frame to the conditioned reconstruction model with the physical computer processor.

Abstract: Systems and methods are disclosed for generating a latent space residual. A computer-implemented method may use a computer system that includes non-transient electronic storage, a graphical user interface, and one or more physical computer processors. The computer-implemented method may include: obtaining a target frame, obtaining a reconstructed frame, encoding the target frame into a latent space to generate a latent space target frame, encoding the reconstructed frame into the latent space to generate a latent space reconstructed frame, and generating a latent space residual based on the latent space target frame and the latent space reconstructed frame.

Abstract: A data processing system includes a computing platform having a hardware processor and a memory storing a data compression software code. The hardware processor executes the data compression software code to receive a series of compression input data and encode a first compression input data of the series to a latent space representation of the first compression input data. The data compression software code further decodes the latent space representation to produce an input space representation of the first compression input data corresponding to the latent space representation, and generates f refined latent values for re-encoding the first compression input data based on a comparison of the first compression input data with its input space representation. The data compression software code then re-encodes the first compression input data using the refined latent values to produce a first compressed data corresponding to the first compression input data.

Abstract: Implementations of the disclosure describe AI systems that offer an improvisational story telling AI agent that may interact collaboratively with a user. In one implementation, a story telling device may be implemented using i) a natural language understanding (NLU) component to process human language input (e.g., digitized speech or text input); ii) a natural language processing (NLP) component to parse the human language input into a story segment or sequence; iii) a component for storing/recording the story as it is created by collaboration; iv) a component for generating AI-suggested story elements; and v) a natural language generation (NLG) component to transform the AI-generated story segment into natural language that may be presented to the user.

Abstract: Systems and methods for a computer-based, interactive communications system capable of generating a response to a human language input are provided. The computer-based, interactive communications systems includes a plurality of response models that may be selected to process one or more keywords extracted from the human language input. The plurality of response models may include at least one conversational response model and at least one informational response model, so that the computer-based, interactive communications system is able to respond to the human language input in a manner commensurate with the type of human language input it receives.

Abstract: Systems and methods for a computer-based, interactive communications system capable of generating a response to a human language input are provided. The computer-based, interactive communications systems includes a plurality of response models that may be selected to process one or more keywords extracted from the human language input. The plurality of response models may include at least one conversational response model and at least one informational response model, so that the computer-based, interactive communications system is able to respond to the human language input in a manner commensurate with the type of human language input it receives.