Abstract: The present disclosure provides for customizable content moderation using neural networks with fine-grained and dynamic image classification ontology. A content moderation system of the present disclosure may provide a plurality of image categories in which a subset of of image categories may be designated as restricted categories. The restricted categories may be chosen by a content provider or an end-user. The content moderation system may utilize a neural network to classify image data (e.g., still images, video, etc.) into one or more of the plurality of image categories, and determine that an image is a restricted image upon classifying the image into one of the restricted categories. The restricted image may by flagged, rejected, removed, or otherwise filtered upon being classified as a restricted image.

Abstract: A method and system for capturing an image of a portion of an environment including an object of interest and uploading of the image to a service for further image analysis. The system captures a first image of an environment or scene using a first sensor of an image capture device. A region of interest is detected within the first image using an object detection model. A set of coordinates corresponding to the portion of the first image is identified and used to position a second sensor. The second sensor captures a second image including the targeted region of interest, wherein the second image has a higher resolution than the first image. The second image is uploaded to an object detection service for the further image analysis.

Abstract: Techniques are described for automatically scaling (or “auto scaling”) compute resources—for example, virtual machine (VM) instances, containers, or standalone servers—used to support execution of service-oriented software applications and other types of applications that may process heterogeneous workloads. The resource requirements for a software application can be approximated by measuring “worker pool” utilization of instances of each service, where a worker pool represents a number of requests that the service can process concurrently. A scaling service can thus be configured to scale the compute instances provisioned for a service in proportion to worker pool utilization, that is, compute instances can be added as the fleet's worker pools become more “busy,” while compute instances can be removed when worker pools become inactive.

Abstract: Techniques for analyzing stored video upon a request are described. For example, a method of receiving a first application programming interface (API) request to analyze a stored video, the API request to include a location of the stored video and at least one analysis action to perform on the stored video; accessing the location of the stored video to retrieve the stored video; segmenting the accessed video into chunks; processing each chunk with a chunk processor to perform the at least one analysis action, each chunk processor to utilize at least one machine learning model in performing the at least one analysis action; joining the results of the processing of each chunk to generate a final result; storing the final result; and providing the final result to a requestor in response to a second API request is described.

Abstract: Approaches are described for a client (e.g., an application executing on a computing device) requiring a particular storage format to provide information describing its storage format to a service, such as a multi-tenant computing environment. The service contains data that the client desires to acquire, but in various embodiments, the data on the service is in a different storage format than the format required by the client. The service receives the information describing the storage format and processes the data accordingly, and then provides the processed data to the client in the format required by the client.

Abstract: Techniques for analyzing stored video upon a request are described. For example, a method of receiving a first application programming interface (API) request to analyze a stored video, the API request to include a location of the stored video and at least one analysis action to perform on the stored video; accessing the location of the stored video to retrieve the stored video; segmenting the accessed video into chunks; processing each chunk with a chunk processor to perform the at least one analysis action, each chunk processor to utilize at least one machine learning model in performing the at least one analysis action; joining the results of the processing of each chunk to generate a final result; storing the final result; and providing the final result to a requestor in response to a second API request is described.

Abstract: A technology is described for managing a temperature of a media device. An example method may include obtaining a current temperature of a media device configured to execute a media application and historical temperature data for the media device for a defined time period. A processing workload strategy may be calculated for the media application that causes a temperature of the media device to stay within a target temperature range for the media device using in part the target temperature range for the media device and the current temperature of the media device to determine a target temperature error, and the historical temperature data for the media device to determine a rate in change of the processing workload strategy based in part on the target temperature error that contains the temperature of the media device within the target temperature range. The media application may then be instructed to manage execution of media using the processing workload strategy.