Abstract: Techniques for anomaly detection are described. An exemplary method includes receiving one or more requests to train an anomaly detection machine learning model using feedback-based training, the request to indicate one or more of a type of analysis to perform, a model selection indication, and a configuration for a training dataset; training the anomaly detection machine learning model according to the one or more requests using the training data; performing feedback-based training on the trained anomaly detection machine learning model; and using the retrained anomaly detection machine learning model.

Abstract: A machine learning environment utilizing training data generated by customer environments. A reinforced learning machine learning environment receives and processes training data generated by independently hosted, or decoupled, customer environments. The reinforced learning machine learning environment corresponds to machine learning clusters that receive and process training data sets provided by the decoupled customer environments. The customer environments include an agent process that collects training data and forwards the training data to the machine learning clusters without exposing the customer environment. The machine learning clusters can be configured in a manner to automatically process the training data without requiring additional user inputs or controls to configured the application of the reinforced learning machine learning processes.

Abstract: Compression profiles may be searched for trained neural networks. An iterative compression profile search may be performed response to a search request. Different prospective compression profiles may be generated for trained neural networks according to a search policy. Performance of compressed versions of the trained neural networks according to the compression profiles may be tracked. The search policy may be updated according to an evaluation of the performance of the compression profiles for the compressed versions of the trained neural networks using compression performance criteria. When a search criteria is satisfied, a result for the compression profile search may be provided.

Abstract: Neural networks with similar architectures may be compressed using shared compression profiles. A request to compress a trained neural network may be received and an architecture of the neural network identified. The identified architecture may be compared with the different network architectures mapped to compression profiles to select a compression profile for the neural network. The compression profile may be applied to remove features of the neural network to generate a compressed version of the neural network.

Abstract: Compression profiles may be searched for trained neural networks. An iterative compression profile search may be performed response to a search request. Different prospective compression profiles may be generated for trained neural networks according to a search policy. Performance of compressed versions of the trained neural networks according to the compression profiles may be tracked. The search policy may be updated according to an evaluation of the performance of the compression profiles for the compressed versions of the trained neural networks using compression performance criteria. When a search criteria is satisfied, a result for the compression profile search may be provided.

Abstract: A compression policy to produce compression profiles for compressing trained machine learning models may be trained using reinforcement learning. An iterative reinforcement learning may be performed response to a search request. Different prospective compression profiles may be generated for received machine learning models according to a compression policy being trained. Performance of compressed versions of the trained neural networks according to the compression profiles may be caused using data sets used to train the machine learning models. The compression policy may be updated according to reward signal determined from an application of a reward function for performance criteria to performance results of the different versions of the machine learning models. When a search criteria is satisfied, the trained compression policy may be provided.

Abstract: Techniques for anomaly detection are described. An exemplary method includes receiving one or more requests to train an anomaly detection machine learning model using feedback-based training, the request to indicate one or more of a type of analysis to perform, a model selection indication, and a configuration for a training dataset; training the anomaly detection machine learning model according to the one or more requests using the training data; performing feedback-based training on the trained anomaly detection machine learning model; and using the retrained anomaly detection machine learning model.

Abstract: Techniques for feedback-based training are described.

Abstract: Implementations detailed herein include description of a computer-implemented method. In an implementation, the computer-implemented method including training a machine learning model using domain mapped third party data; and performing inference using the machine learning model by: receiving scoring data, domain mapping the received scoring data using a domain mapper that was used to generate the domain mapped third party data, and applying the machine learning model to the domain mapped received scoring data to generate an output result.

Abstract: Features related to systems and methods expediting generation of a machine learning model, such as an image recognition model, are described. Existing machine learning models are analyzed to identify a starting point for creating the new machine learning model. An existing machine learning model can suggest learning parameters (e.g., training parameters or structural features of the model) that can be used to expedite the generating and training process along with training data that can augment the training of the new machine learning model.

Abstract: A method and system for a deep learning-based approach to image processing to increase a level of optical zooming and increasing the resolution associated with a captured image. The system includes an image capture device to generate a display of a field of view (e.g., of a scene within a viewable range of a lens of the image capture device). An indication of a desired zoom level (e.g., 1.1× to 5×) is received, and, based on this selection, a portion of the field of view is cropped. In one embodiment, the cropped portion displayed by the image capture device for a user's inspection, prior to the capturing of a low resolution image. The low resolution image is provided to an artificial neural network trained to apply a resolution up-scaling model to transform the low resolution image to a high resolution image of the cropped portion.

Abstract: Features related to systems and methods for generating a machine learning model that is a composite of at least two other models (e.g., crowd-sourced models contributed by users) are described. Each of the contributed models provide output values that may not be to scale. To account for these differences, a normalization factor for a first machine learning model is generated to adjust values produced by the first machine learning model to correspond with results from the second machine learning model. The crowd-sourced models along with the normalization factor are included in the new image model generated in the claims.

Abstract: Systems and methods for detecting motion in compressed video are provided. Some methods can include parsing a stream of compressed video, obtaining macroblock size information from the parsed stream, computing factors derived from the macroblock size information, computing adaptive threshold values derived from relative frame characteristics of the compressed video, comparing the factors derived from the macroblock size information with the adaptive threshold values, and detecting motion based upon the comparing when at least one of the factors exceeds at least one of the adaptive threshold values. In some embodiments, detecting the motion can include performing spatio-temporal filtering on macroblocks in which the motion is detected or performing spatio-temporal filtering on at least one non-motion macroblock.

Abstract: Disclosed are various embodiments for improving optical character recognition approaches through the use of gap shifting. A text detection process is performed upon an image to detect a first region of text. A second region that is in line with the first region is shifted to reduce a gap between the first region and the second region, thereby creating a modified image. The text detection process is performed upon the modified image in order to detect text within the second region.

Abstract: Systems and methods of motion detection on encrypted or scrambled video data streams are provided. Some methods can include identifying macroblock size information for an encrypted/scrambled video data stream and using the identified macroblock size information to determine a presence of motion in the encrypted/scrambled video data stream without decrypting and descrambling the encrypted/scrambled video data stream.

Abstract: Systems and methods for detecting motion in compressed video are provided. Some methods can include parsing a stream of compressed video, obtaining macroblock size information from the parsed stream, computing factors derived from the macroblock size information, computing adaptive threshold values derived from relative frame characteristics of the compressed video, comparing the factors derived from the macroblock size information with the adaptive threshold values, and detecting motion based upon the comparing when at least one of the factors exceeds at least one of the adaptive threshold values. In some embodiments, detecting the motion can include performing spatio-temporal filtering on macroblocks in which the motion is detected or performing spatio-temporal filtering on at least one non-motion macroblock.

Abstract: A method and apparatus wherein the method includes the steps of parsing a stream of compressed video, obtaining macroblock size information from the parsed stream, computing factors derived from the macroblock size, wherein the factors include a normalized bit size, a bit size ratio and a neighbor score, computing corresponding adaptive threshold values derived from the relative frame characteristics of the compressed video, comparing the factors derived from the macroblock size information with the corresponding adaptive threshold values and detecting motion based upon combinations of the comparisons when the factors exceed the threshold value.

Abstract: Devices, methods, and systems for determining pushback direction are described herein. One method includes receiving a video image of an aircraft during a pushback of the aircraft, determining a motion flow map associated with the video image during the pushback, determining a motion orientation histogram associated with the video image during the pushback, and determining a direction of the pushback based on the motion flow map and the motion orientation histogram.

Abstract: A method includes receiving an image of a face to match with images of known faces, extracting blocks multiple blocks from the received image, calculating local binary pattern histograms for each block, generating matching scores for each block against block of the images of known faces, determining a top number, N, of matching scores less than the number of blocks, and matching the received image to an image of a known face as a function of the top number of matching scores.

Abstract: A system and method include obtaining an image of an analog dial gauge. The image is processed to identify an endpoint of the gauge and a needle position in the image. A reading of the gauge is determined from the endpoint, the needle position, and information regarding the range of the gauge.