Abstract: In various examples, map data or geospatial data is used to identify a subset of sensor data having a higher likelihood of including representations of a target object of interest from a larger set of sensor data. Feature vectors corresponding to the subset of sensor data may then be compared to template feature vectors corresponding to the target object in order to confirm the depiction of the target object in the sensor data. The identified sensor data may be used to train one or more machine learning model to compute outputs that correspond to object identification. The trained machine learning models may be used to identify objects in order to aid an autonomous or semi-autonomous machine navigating within a surrounding environment.

Abstract: In various examples, surface profile estimation and bump detection may be performed based on a three-dimensional (3D) point cloud. The 3D point cloud may be filtered in view of a portion of an environment including drivable free-space, and within a threshold height to factor out other objects or obstacles other than a driving surface and protuberances thereon. The 3D point cloud may be analyzed—e.g., using a sliding window of bounding shapes along a longitudinal or other heading direction—to determine one-dimensional (1D) signal profiles corresponding to heights along the driving surface. The profile itself may be used by a vehicle—e.g., an autonomous or semi-autonomous vehicle—to help in navigating the environment, and/or the profile may be used to detect bumps, humps, and/or other protuberances along the driving surface, in addition to a location, orientation, and geometry thereof.

Abstract: The present disclosure provides an adaptor for managing data flow in an inference pipeline. The adaptor comprises a non-transitory memory device storing instruction code modules and at least one processor configured to execute the modules. The processor is configured to collect model metadata describing at least one machine-learning (ML) based inference model deployed within the inference pipeline, define a required ingress data format for interfacing the inference model based on the model metadata, employ an ingress adaptation module to receive an input data element from one or more data input sources and translate the input data element according to the ingress data format, and process the translated input data element through the inference model to generate at least one first inference result.

Abstract: In various examples, accuracy determinations for localization in autonomous and semi-autonomous systems and applications are described herein. Systems and methods are disclosed that determine one or more errors associated with vehicle localization using various types of sensor data generated using a vehicle. For instance, a first component of the vehicle may use a map and first sensor data to determine an estimated pose of the vehicle. A second component of the vehicle may then determine the error(s) associated with the estimated pose based on both actual motion of the vehicle within the environment, as determined using second sensor data, and comparing features represented by the first sensor data to features represented by the map. In some examples, the second component may further determine information associated with the error(s), such as one or more uncertainties associated with the error(s).

Abstract: Various examples, systems, and methods are disclosed relating to a content curation pipeline. A first computing system obtain, from at least one data source, a plurality of content records corresponding with at least one content source of the at least one data source. The first computing system determine a subset of a plurality of embeddings corresponding to at least one category. The first computing system apply the subset of the plurality of embeddings as input to at least one language model (LM) to cause the at least one LM to generate, using a template, at least one structured output comprising a plurality of modified content and corresponding metadata. The first computing system provide, via an interface, the at least one structured output.

Abstract: In various examples, images (e.g., novel views) of an object may be rendered using an optimized number of samples of a 3D representation of the object. The optimized number of the samples may be determined based at least on casting rays into a scene that includes the 3D representation of the object and/or an acceleration data structure corresponding to the object. The acceleration data structure may include features corresponding to characteristics of the object, and the features may be indicative of the number of samples to be obtained from various portions of the 3D representation of the object to render the images. In some examples, the 3D representation may be a neural radiance field that includes, as a neural output, a spatially varying kernel size predicting the characteristics of the object, and the features of the acceleration data structure may be related to the spatially varying kernel size.

Abstract: In various examples, systems and methods are disclosed relating to registering image processing with image registration for image generation and content stream applications. Systems and methods are disclosed for registering portions of images that are modified to incorporate content or features, with references images from which the portions of the images are identified. The systems and methods can transform the modified portions to more realistically and precisely merge back into the reference images, such as for presentation as a content stream.

Abstract: Disclosed are apparatuses, systems, and techniques that reduce latency of frame processing pipelines. The techniques include but are not limited to causing a display to set a refresh rate that matches a frame rendering rate of an application and rendering, with the frame rendering rate, a plurality of frames. Rendering frames includes generating, using a first processing unit, sets of instructions associated with respective frames and generated starting at times spaced with the refresh rate, processing, using a second processing unit, the sets of instructions to render the frames, and causing the display to display the rendered frames.

Abstract: Apparatuses, systems, and techniques to generate gaming strategies using neural networks. In at least one embodiment, one or more neural networks are used to generate one or more gaming strategy outputs based, at least in part, on one or more text inputs describing one or more multimedia gaming contents.

Abstract: In various examples, machine learning model-based sign language corrective feedback is provided. Machine learning model-based recognition of sign language symbols (e.g., body poses and/or movements) may be used to generate real-time kinematic feedback to a sign language speaker that guides them to adjust their hand pose to correctly align with established sign language standard symbols. A sign language feedback framework may process video data representing conversational sign language symbol sequences to extract kinematic keypoints to search a sign language dictionary representing an established vocabulary of sign language symbols. Based on selecting an intended sign language symbol from the dictionary and computing deviations between the kinematic keypoint pattern of the intended sign language symbol from the dictionary and the extracted kinematic keypoint symbol, the sign language feedback framework may provide real-time kinematic feedback to the signer describing how to adjust their signing.

Abstract: Apparatuses, systems, and techniques to perform an application programming interface (API) to cause dependency type information of one or more graph nodes of a software graph to be modified. In at least one embodiment, one or more dependency types from a graph are modified.

Abstract: Apparatuses, systems, and techniques to format memory to store values of mixed precision. In at least one embodiment, processor(s) perform operations to format memory to store values of mixed precision in a block of contiguous memory by, in part, using an indication of a number of higher-precision values to be stored in memory.

Abstract: Systems and methods for atomic memory operations in a remote direct memory access network are disclosed. A system includes a network interface card (NIC) comprising a first memory and one or more processors coupled to the first memory. The one or more processors are to receive an atomic memory operation (AMO) remote procedure call (RPC) comprising a memory address and an AMO type. The one or more processors are further to retrieve a value corresponding to the memory address of the AMO RPC from a second memory. The one or more processors are further to perform an AMO corresponding to the AMO type on the value from the second memory to obtain a modified value. The one or more processors are further to store the modified value in the first memory.

Abstract: In various examples, a technique for performing conditional data sourcing and curation includes retrieving, by a plurality of processing nodes, a plurality of content items from one or more data sources, wherein each processing node retrieves a different subset of the plurality of content items from the one or more data sources. The technique also includes applying a first set of filters to metadata associated with the content items to generate a plurality of filtered content items. The technique further includes generating, based on a subset of the metadata associated with the filtered content items and a second set of filters, mappings between the filtered content items and text descriptions for the filtered content items and storing, based on the mappings, the filtered content items in association with the text descriptions in one or more data stores.

Abstract: Traditionally, a software application is developed, tested, and then published for use by end users. Any subsequent update made to the software application is generally in the form of a human programmed modification made to the code in the software application itself, and further only becomes usable once tested, published, and installed by end users having the previous version of the software application. This typical software application lifecycle causes delays in not only generating improvements to software applications, but also to those improvements being made accessible to end users. To help avoid these delays and improve performance of software applications, deep learning models may be made accessible to the software applications for use in providing inferenced data to the software applications, which the software applications may then use as desired. These deep learning models can furthermore be improved independently of the software applications using manual and/or automated processes.

Abstract: Apparatuses, systems, and techniques for testing processing units of a computing system are disclosed herein. A request to initiate a testing process for each of a set of processing units is received. A first processing unit includes a first virtual processor executing first operations. A second processing unit includes a second virtual processor executing second operations. Execution of the first operations is transferred from the first virtual processor to the second virtual processor. Execution of the testing process is initiated at the first processing unit while the second virtual processor executes the first and second operations. In response to a detection that the execution of the testing process is completed, execution of the first and second operations is transferred to the first virtual processor. Execution of the testing process is initiated at the second processing unit while the first virtual processor running executes the first and second operations.

Abstract: In various examples, systems and methods are disclosed that perform motion detection across image frames, such as optical flow determination, to synchronize an asynchronous frame with respect to a target time for the asynchronous frame. For example, image frames from a sensor can be processed by an optical flow accelerator to detect displacement across the image frames, and the displacement can be used to interpolate a modified frame at the target time. This can be used to perform data collection and combining operations such as stitching and/or reconstruction. The synchronization can be performed from sensor data from sensors such as cameras, LIDAR sensors, and/or RADAR sensors.

Abstract: Apparatuses, systems, and techniques for calculating data compression parameters using codebook entry values. In at least one embodiment, one or more circuits is to calculate one or more data compression parameters based, at least in part, on at least on one or more values of the data to be compressed in relation to at least two codebook entry values.

Abstract: In various examples, image-based three-dimensional occupant assessment for in-cabin monitoring systems and applications are provided. An evaluation function may determine a 3D representation of an occupant of a machine by evaluating sensor data comprising an image frame from an optical image sensor. The 3D representation may comprise at least one characteristic representative of a size of the occupant, (e.g., a 3D pose and/or 3D shape), which may be used to derive other characteristics such as, but not limited to weight, height, and/or age. A first processing path may generate a representation of one or more features corresponding to at least a portion of the occupant based on optical image data, and a second processing path may determine a depth corresponding to the one or more features based on depth data derived from the optical image data and ground truth depth data corresponding to the interior of the machine.

Abstract: Disclosed are systems and techniques for training personalized language models. The techniques include applying a plurality of first machine learning models to a first input prompt. Each of the plurality of first machine learning models generates a respective reward value of a first plurality of reward values. The techniques include applying a second machine learning model to the first plurality of reward values to obtain first reward value embeddings; applying a third machine learning model to the first reward value embeddings and the first input prompt to obtain a first output response; calculating a first loss based on a comparison between the first output response and the first input prompt; and causing the second machine learning model to be modified based on the first loss.