Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining routing. An exemplary method comprises: inputting a plurality of to-be-optimized routing solution candidates to a Siamese neural network comprising a plurality of value prediction networks, each of the value prediction networks being trained to predict a cost associated with a to-be-optimized routing solution candidate; identifying one or more to-be-optimized routing solution candidates from the plurality of to-be-optimized routing solution candidates based on outputs of the Siamese neural network; inputting the one or more identified to-be-optimized routing solution candidates to a routing optimizer to obtain one or more optimized routing solution candidates; and determining an optimized routing solution with a lowest cost from the one or more optimized routing solution candidates.

Abstract: Implementations are directed to training a machine learning model that, once trained, is used in performance of robotic grasping and/or other manipulation task(s) by a robot. The model can be trained using simulated training examples that are based on simulated data that is based on simulated robot(s) attempting simulated manipulations of various simulated objects. At least portions of the model can also be trained based on real training examples that are based on data from real-world physical robots attempting manipulations of various objects. The simulated training examples can be utilized to train the model to predict an output that can be utilized in a particular task—and the real training examples used to adapt at least a portion of the model to the real-world domain can be tailored to a distinct task. In some implementations, domain-adversarial similarity losses are determined during training, and utilized to regularize at least portion(s) of the model.

Abstract: A system and method for training a computer-implemented object classifier includes detecting a foreground visual object within a sub-region of a scene, determining a background model of the sub-region of the scene, the background model representing the sub-region when any foreground visual object is absent from that sub-region, and training the object classifier by computer-implemented machine learning using the background model of the sub-region as a negative training example.

Abstract: A method and apparatus for removing black eyepits and sunglasses in first actual scenario data having an image containing a face acquired from an actual scenario, to obtain second actual scenario data; counting a proportion of wearing glasses in the second actual scenario data; dividing original training data composed of an image containing a face into wearing-glasses and not-wearing-glasses first and second training data, where a proportion of wearing glasses in the original training data is lower than a proportion in the second actual scenario data; generating wearing-glasses third training data based on glasses data and the second training data; generating fourth training data in which a proportion of wearing glasses is equal to the proportion of wearing glasses in the second actual scenario data, based on the third training data and the original training data; and training a face recognition model based on the fourth training data.

Abstract: A computerized automated parameterization image pattern detection and classification method performs (1) morphological metrics learning using labeled region data to generate morphological metrics; (2) intensity metrics learning using learning image and labeled region data to generate intensity metrics; and (3) population learning using the morphological metrics and the intensity metrics to generate learned pattern detection parameter. The method may further update the learned pattern detection parameter using additional labeled region data and learning image, and apply pattern detection with optional user parameter adjustment to image data to generate detected pattern. The method may alternatively perform pixel parameter learning and pixel classification to generate pixel class confidence, and uses the pixel class confidence and the labeled region data to perform pattern parameter learning to generate the learned pattern detection parameter.

Abstract: A moving object detection apparatus has a generation unit that generates a background model based on a feature of a background region of a captured image that is captured by image capturing unit, a detection unit that detects a moving object region from an image input using an input unit, based on the background model, and a determination unit that determines whether to cause the generation unit to newly generate a background model, based on an amount of change in the moving object region detected by the detection unit for a first image and a second image captured at different times and input using the input unit.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing inputs using an image processing neural network system that includes a spatial transformer module. One of the methods includes receiving an input feature map derived from the one or more input images, and applying a spatial transformation to the input feature map to generate a transformed feature map, comprising: processing the input feature map to generate spatial transformation parameters for the spatial transformation, and sampling from the input feature map in accordance with the spatial transformation parameters to generate the transformed feature map.

Abstract: Provided are a method for pushing a picture, a terminal and a storage medium. The method includes the follows. An image containing a facial expression of a user is acquired, when a screen is lit; the image is analyzed to determine an emotion type of the user; and a target picture is pushed according to the emotion type of the user.

Abstract: Techniques are disclosed for improving classification performance in supervised learning. In accordance with some embodiments, a multiclass support vector machine (SVM) having three or more classes may be converted to a plurality of binary problems that then may be reduced via one or more reduced-set methods. The resultant reduced-set (RS) vectors may be combined together in one or more joint lists, along with the original support vectors (SVs) of the different binary classes. Each binary problem may be re-trained using the joint list(s) by applying a reduction factor (RF) parameter to reduce the total quantity of RS vectors. In re-training, different kernel methods can be combined, in accordance with some embodiments. Reduction may be performed until desired classification performance is achieved. The disclosed techniques can be used, for example, to improve classification speed, accuracy, class prioritization, or a combination thereof, in the SVM training phase, in accordance with some embodiments.

Abstract: A system for generating graphical user interfaces. The system may include processors and storage devices storing instructions. The instructions may configure the one or more processors to perform operations including identifying a plurality of attributes from an image captured with a client device, identifying a plurality of first results based on the attributes, generating a first graphical user interface for display in the client device. The first graphical user interface may include a plurality of result icons corresponding to a subset of the first results having confidence scores above a threshold, a plurality of filter icons displaying options, and a search button. The operations may also include receiving a selection of at least one of the result icons or at least one of the filter icons, performing a search, based on the selection, and generating a second graphical user interface.

Abstract: Various arrangements for generating a high definition map are presented. An image of a roadway environment may be captured. A laser imaging detection and ranging point cloud of the roadway environment may be generated. An object recognition process may be performed on the image of the roadway environment to detect one or more objects present in the roadway environment. A fusion process may be performed using the lidar point cloud to identify a location of the one or more detected objects from the image in a vehicle frame of reference. The one or more objects may be mapped from the vehicle frame of reference to a global frame of reference An autonomous driving high definition map may be created that includes the mapped one or more objects in the global frame of reference.

Abstract: Hybrid transcription of audio relies on having one or more layers of transcribers who review transcriptions generated by automatic speech recognition (ASR) systems in order to correct errors that are found in the transcriptions. When it comes to determining how much human reviewing is needed, such as determining how many layers of review to use, there is a cost/benefit tradeoff to consider. Some embodiments described herein utilize a machine learning-based approach for estimating quality of hybrid transcription of audio. In one embodiment, a computer generates a transcription of a segment of audio using an ASR system, which is subsequently reviewed by a transcriber. The computer then calculates, based on properties of the review by the transcriber, a value indicative of an expected accuracy of the reviewed transcription. The computer may suggest a second transcriber review the reviewed transcription if the value indicative of the expected accuracy is below a threshold.

Abstract: A lighting unit for use in low lighting environments such as underground mines is provided. The lighting unit includes a main housing having a front opening and a cavity, a first visual lighting assembly positioned within the cavity of the main housing that emits light in the visible spectrum, and a second ultraviolet lighting assembly positioned a fixed distance from the first visual lighting assembly within the cavity of the main housing that emits light in the ultraviolet spectrum. The lighting unit is part of a system for detecting patterns on objects having reflective or high visibility colored material thereon. In addition to the lighting unit, the system includes a camera having a field of view within which the system is configured to detect patterns of reflective or high visibility colored material on one or more objects, a warning device to alert the operator when a marked object enters the camera's field of view.

Abstract: A system that includes a first device configured to implement a first sensor engine in signal communication with a second device configured to implement a first actor engine. The first sensor engine is configured to receive a first real world value, to identify a real world value in a first sensor table based on the first real world value, to fetch a first correlithm object linked with the identified real world value entry from the first sensor table, and to send the first correlithm object to a second device using the network connection. The first actor engine is configured to receive the first correlithm object, to identify a correlithm object from a first actor table with the shortest distance, to fetch a second real world value from the first actor table linked with the identified correlithm object, and to output the second real word value.

Abstract: Systems and methods for generating biomarkers associated with neuropsychiatric disorders, neurodevelopmental disorders, neurobehavioral disorders, or other neurological disorders are described. In general, the biomarkers are generated based on correlations between functional imaging data and clinical acquired from a subject, as computed using a multivariate classifier. Functional imaging data may include functional magnetic resonance images, or activation maps generated from such images. Clinical data generally includes data associated with a clinical or behavioral characterization of the subject. The biomarkers can be used to monitor or otherwise assess a treatment response; to provide diagnostic information, such as subtyping or classifying a disorder; to provide prognostic information, such as a prediction of treatment response or outcome; or to indicate functional or anatomical targets for treatments.

Abstract: Methods and systems disclosed herein may quantify the content and nature of a first stream of text to detect when the typical composition of the first stream of text changes. Quantifying the content and nature of the first stream of text may begin by generating a baseline representation of the content of the first stream of text as represented by a first matrix. Once generated, the first matrix may be used as a control against subsequently received sequences of text. In this regard, a second matrix may be generated from a second sequence of text and compared to the first matrix to determine the differences between the first sequence of text and the second sequence of text. Once a difference is determined, the difference may be compared to a threshold value and, when the difference exceeds the threshold value, an administrator may be notified and corrective action taken.

Abstract: A method for tracking workers at a worksite is disclosed. The method includes acquiring a video of the worksite and detecting a movement of objects in the video. Each object is represented by a plurality of pixels. The method includes identifying an item associated with at least one object by clustering a set of pixels of the plurality of pixels based on a color of the set of pixels. Next, the method includes determining the item as a hard hat based on at least a shape of the set of pixels, and identifying the object as a specific worker based on the color of the set of pixels. The method also includes generating a movement data associated with the specific worker based on a tracking of the hard hat.

Abstract: A server receives an input image group from an MFP. The server decides a reference image from the input image group, and generates rotated images by deforming the reference image. The server calculates image feature amounts from the reference image and the rotated images, and generates direction determiners using the image feature amounts. With respect to each of images of the input image group, the server determines the directions of the images by applying the direction determiners, and when the directions of the images are different from the direction of the reference image, the images are rotated to the direction of the reference image, and the directions of each of the images of the input image group are unified. The server generates a classification rule to be used in classification of images using the input image group in which the directions are unified.

Abstract: Methods and apparatuses are described for of recognizing handwritten characters in digital images using context-based machine learning. A server captures an image of a document that comprises one or more handwritten data fields, the document associated with a user identifier. The server identifies a field type for each handwritten data field in the image. The server creates a pixel intensity array for each character in each handwritten data field and determines whether a user-specific character map exists for the user identifier. If a map exists, the server retrieves the map and generates digital form data by executing a user-specific handwriting classifier using the map, the pixel intensity arrays, and the field types. If a map does not exist, the server builds a map based upon the pixel intensity arrays and generates digital form data by executing a baseline handwriting classifier using the map, the pixel intensity arrays, and the field types.

Abstract: In a neural network unit, each neural processing unit (NPU) of an array of N NPUs receives respective first and second upper and lower bytes of 2N bytes received from first and second RAMs. In a first mode, each NPU sign-extends the first upper byte to form a first 16-bit word and performs an arithmetic operation on the first 16-bit word and a second 16-bit word formed by the second upper and lower bytes. In a second mode, each NPU sign-extends the first lower byte to form a third 16-bit word and performs the arithmetic operation on the third 16-bit word and the second 16-bit word formed by the second upper and lower bytes. In a third mode, each NPU performs the arithmetic operation on a fourth 16-bit word formed by the first upper and lower bytes and the second 16-bit word formed by the second upper and lower bytes.

Abstract: A method for economizing computing resources and verifying an integrity of parameters of a neural network by inserting test pattern into a background area of an input image is provided for fault tolerance, fluctuation robustness in extreme situations, functional safety on the neural network, and an annotation cost reduction. The method includes: a computing device (a) generating t-th background prediction information of a t-th image by referring to information on each of a (t?2)-th image and a (t?1)-th image; (b) inserting the test pattern into the t-th image by referring to the t-th background prediction information, to thereby generate an input for verification; (c) generating an output for verification from the input for verification; and (d) determining the integrity of the neural network by referring to the output for verification and an output for reference. According to the method, a data compression and a computation reduction are achieved.

Abstract: An online system ranks test cases run in connection with check-in of sets of software files in a software repository. The online system ranks the test cases higher if they are more likely to fail as a result of defects in the set of files being checked in. Accordingly, the online system informs software developers of potential defects in the files being checked in early without having to run the complete suite of test cases. The online system determines a vector representation of the files and test cases based on a neural network. The online system determines an aggregate vector representation of the set of files. The online system determines a measure of similarity between the test cases and the aggregate vector representation of the set of files. The online system ranks the test cases based on the measures of similarity of the test cases.

Abstract: Methods and systems involving convolutional neural networks as applicable for semantic segmentation, including multi-task convolutional networks employing curriculum based transfer learning, are disclosed herein. In one example embodiment, a method of semantic segmentation involving a convolutional neural network includes training and applying the convolutional neural network. The training of the convolutional neural network includes each of training a semantic segmentation decoder network of the convolutional neural network, generating first feature maps by way of an encoder network of the convolutional neural network, based at least in part upon a dataset received at the encoder network, and training an instance segmentation decoder network of the convolutional neural network based at least in part upon the first feature maps.

Abstract: An image processing apparatus includes a first learning unit configured to learn an identifier for identifying a class of a region formed by segmenting an image based on first training data, an evaluation unit configured to evaluate a result of identification of a class of the first training data by the identifier, a generation unit configured to generate second training data from the first training data based on an evaluation result by the evaluation unit, and a second learning unit configured to learn a plurality of identifiers different from the identifier learned by the first learning unit based on the second training data.

Abstract: Multiple 3×3 convolutional filter kernels are used for approximating operations of fully-connected (FC) layers. Image classification task is entirely performed within a CNN based integrated circuit. Output at the end of ordered convolutional layers contains P feature maps with F×F pixels of data per feature map. 3×3 filter kernels comprises L layers with each organized in an array of R×Q of 3×3 filter kernels, Q and R are respective numbers of input and output feature maps of a particular layer of the L layers. Each input feature map of the particular layer comprises F×F pixels of data with one-pixel padding added around its perimeter. Each output feature map of the particular layer comprises (F?2)×(F?2) pixels of useful data. Output of the last layer of the L layers contains Z classes. L equals to (F?1)/2 if F is an odd number. P, F, Q, R and Z are positive integers.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using neural networks. One of the methods includes receiving, by a neural network in a speech recognition system, first data representing a first raw audio signal and second data representing a second raw audio signal, the first raw audio signal and the second raw audio signal for the same period of time, generating, by a spatial filtering convolutional layer in the neural network, a spatial filtered output the first data and the second data, generating, by a spectral filtering convolutional layer in the neural network, a spectral filtered output using the spatial filtered output, and processing, by one or more additional layers in the neural network, the spectral filtered output to predict sub-word units encoded in both the first raw audio signal and the second raw audio signal.

Abstract: Grain noise and scratches are applied to an input image as random noise to generate a combined image. When applying grain noise clipped from two-dimensional noise data to the input image, each time clipping is performed, a positional difference from a last clipping position is evaluated. When the positional difference is determined to be small, a current clipping position is changed. When applying scratches clipped from noise data of a plurality of patterns to an input image, when a condition for successively applying scratches for a predetermined time period is selected, a positional difference from a last pasting position is evaluated. When the positional difference is determined to be large, the application of scratches is invalidated.

Abstract: A neuromorphic electric system includes a network of plural neuron circuits connected in series and in parallel to form plural layers. Each of the plural neuron circuits includes: a soma circuit that stores a charge supplied thereto and outputs a spike signal; and plural synapse circuits that supply a charge to the soma circuit according to a spike signal fed to the synapse circuits, a number of the plural synapse circuits being one more than a number of plural neuron circuits in a prior layer outputting the spike signal to the synapse circuits. One of the plural synapse circuits supplies a charge to the soma circuit in response to receiving a series of pulse signals, and the others of the plural synapse circuits supply a charge to the soma circuit in response to receiving a spike signal from corresponding neuron circuits in the prior layer.

Abstract: A network device, system, and method are provided. The network device includes a processor. The processor is configured to store a local estimate and a dual variable maintaining an accumulated subgradient for the network device. The processor is further configured to collect values of the dual variable of neighboring network devices. The processor is also configured to form a convex combination with equal weight from the collected dual variable of neighboring network devices. The processor is additionally configured to add a most recent local subgradient for the network device, scaled by a scaling factor, to the convex combination to obtain an updated dual variable. The processor is further configured to update the local estimate by projecting the updated dual variable to a primal space.

Abstract: Machine training and application of machine-trained classifier are used for image-based tumor phenotyping in a medical system. To create a training database with known phenotype information, synthetic medical images are created. A computational tumor model creates various examples of tumors in tissue. Using the computational tumor model allows one to create examples not available from actual patients, increasing the number and variance of examples used for machine-learning to predict tumor phenotype. A model of an imaging system generates synthetic images from the examples. The machine-trained classifier is applied to images from actual patients to predict tumor phenotype for that patient based on the knowledge learned from the synthetic images.

Abstract: Described is a system for modeling probability matching in human subjects. Features related to probability matching are extracted from a set of human subject responses from behavioral tasks. Neural network model instances are trained on the set of features, resulting in a set of trained neural network model instances. A set of model parameters are derived from the set of trained neural network instances, and the set of derived model parameters are used to emulate human performance on novel data.

Abstract: A computer-implemented method and a system are proposed. According to the method, in response to receiving a character, a first representation of the character is generated by performing word embedding processing on the character. The first representation is related to context of the character. A second representation of the character is generated by performing convolutional neural network (CNN) processing on the character. The second representation is related to a hieroglyphic feature of the character. A label for the character is determined by performing recurrent neural network (RNN) processing on the first representation and the second representation. The label indicates an attribute of the character related to the context.

Abstract: A computer-implemented method and a system are proposed. According to the method, in response to receiving a character, a first representation of the character is generated by performing word embedding processing on the character. The first representation is related to context of the character. A second representation of the character is generated by performing convolutional neural network (CNN) processing on the character. The second representation is related to a hieroglyphic feature of the character. A label for the character is determined by performing recurrent neural network (RNN) processing on the first representation and the second representation. The label indicates an attribute of the character related to the context.

Abstract: At least some embodiments of the present disclosure feature systems and methods for assessing the impact of visual features within a region of a scene. With the input of a visual representation of a scene and at least one selected region within the scene, the system applies a visual attention model to the visual representation to determine visual conspicuity of the at least one selected region. The system computes feature-related data associated with a plurality of features of the at least one selected region. Based on the visual conspicuity and the feature-related data, the system assesses an impact that at least one of the features within the at least one selected region have on the visual conspicuity.

Abstract: Software testing techniques based on image recognition are disclosed. In various embodiments, a programmatically implemented image classifier is trained to recognize a screen shot image as being associated with a transaction end condition of a transaction. A test script configured to initiate an iteration of the transaction is run. A start time of the iteration of the transaction is recorded. Screen shot images are generated during performance of the iteration of the transaction to capture a series of screen shot images of at least a portion of a user interface display associated with the iteration of the transaction. The image classifier is used to find an earliest-captured image that matches the transaction end condition. A time associated with the matched image is used as a transaction end time to compute an end-to-end time to perform the iteration of the transaction.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for correcting entity names. One method includes receiving texts and deriving a plurality of name-context pairs from the texts. The method further includes calculating a context consistency measure for each name-context pair and storing context-entity name data representing the name-context pairs. Another method includes identifying an entity name and one or more context terms from a query and generating candidate names for the entity name. The method further includes determining a score for each of the candidate names, selecting a number of top scoring candidate names, and using the selected candidate names to respond to the query.

Abstract: There is provided a system configured to receive a plurality of images, analyze a set of features of the plurality of images to determine a difference between a first training performance of a plurality of independent detectors based on one or more of individual attributes and a second training performance of a plurality of joint detectors based on one or more of composite attributes, select, based on the analyzing, either one of the plurality of independent detectors or one of the plurality of joint detectors for identifying a plurality of objects in the plurality of images, and identify the plurality of objects in the plurality of images, using the selected one of the plurality of independent detectors utilizing the one or more of the individual attributes or using the selected one of the plurality of joint detectors utilizing the one or more of the composite attributes.

Abstract: The technology described herein provides an efficient mechanism for generating image tags. Image data from a plurality of sources may be analyzed to identify relevant text items from the aggregated data. The relevant text items may be keywords describing a subject of an image, an entity of an image, a location of an image, or the like. From the aggregated image data, one or more image tags may be generated and stored as an offline dataset with an image identifier. Upon detecting a prompt such as a user issuing a search query for an image, the image identifier is used to perform a look up of the image and associated image tags to be provided.

Abstract: Various examples of methods and systems are provided for approximate message passing with universal denoising. In one example, a method includes applying an approximate message passing (AMP) conversion framework to a plurality of substantially linear measurements to produce a plurality of scalar measurements; applying a denoiser algorithm to the plurality of scalar measurements to generate a plurality of denoised scalar measurements; and applying a conversion term to the plurality of denoised scalar measurements to convert the plurality of denoised scalar measurements to a plurality of denoised substantially linear measurements.

Abstract: In some embodiments, the present invention provides for an exemplary inventive system, including: a communication pipeline, including: at a first end of the communication pipeline: a first processor configured to: obtain a plurality of original content data units having a representative content associated with a subject; apply a trained artificial intelligence algorithm to identify: the representative content of the subject and original background content that is not associated with the subject; remove the original background content to reduce a volume of data being transmitted resulting in an increased capacity of the communication channel; encode and transmit each respective modified content data unit from the first end of the communication pipeline to a second end; a second processor configured to: receive and decode each respective modified content data unit; generate a respective artificial background content; and combine the representative content associated with the subject and the respective artificia

Abstract: A grammatically diverse test set of natural language sentences for a deep question answering system is provided by analyzing a given sentence to characterize its syntactical classification, and adding the sentence to the test set if its classification is sufficiently different from other sentences already in the test set. A particular sentence may be selected for inclusion according to a desired syntactic distribution. Multiple sentences having the exact same classification may be allowed subject to a maximum number of such sentences. The test set is adapted to an element of interest by characterizing each syntactical classification relative to the element of interest. The analysis derives a parse tree, identifies a particular node of the tree corresponding to the element of interest, and extracts syntactic information by traversing the tree starting at the particular node and ending at the root node of the tree according to different traversal schemes.

Abstract: Described are an apparatus and method for generating to generate an acoustic model. The apparatus and method include a processor a processor configured to calculate a noise representation that represents noise data by using a noise model, and generate the acoustic model through training using training noisy speech data, which comprises speech data and the noise data, a string of phonemes corresponding to the speech data, and the noise representation.

Abstract: An image processing apparatus is configured to process an image. The image processing apparatus includes: a calculator configured to calculate a parameter for each of a plurality of areas and for each of a plurality of models describing a plurality of body shapes, each of the plurality of models having the plurality of areas, and probabilities indicating to which of the plurality of models input images input for learning belong; and a comparator configured to compare an input inspection image with the plurality of models whose parameters are optimized based on the probabilities.

Abstract: A system generates a ticker result, which may be a uniform resource locator (URL) corresponding to a quote provider. The system receives a string of information and determines whether all terms in the string of information correspond to ticker symbols. If all terms in the string of information correspond to ticker symbols, the system may ascertain whether the string of information corresponds to a query for ticker information. If the string of information corresponds to a query for ticker information, the ticker information (e.g., a ticker result) may be provided.

Abstract: Achieving high spatial resolution in contact sensing for robotic manipulation often comes at the price of increased complexity in fabrication and integration. One traditional approach is to fabricate a large number of taxels, each delivering an individual, isolated response to a stimulus. In contrast, proposed sensor includes a continuous volume of soft material, e.g., a piezoresistive elastomer with a number of terminals embedded inside. Piezoresistive effects can be measured between all pairs of terminals in the set, and this rich signal set can contain the information needed to pinpoint contact location with high accuracy using regression algorithms. Submillimeter median accuracy can be demonstrated in locating contact on a 10 mm by 16 mm sensor using only four terminals (creating six unique pairs).

Abstract: Efficient image processing systems and methods for image scene classification and similarity matching are disclosed. The image processing systems encompassed by this disclosure use a deep convolutional neural network to facilitate scene classification by recognizing the context of an image and thereby enabling searches for similar images. These methods and systems are scalable to a large set of images and achieve a higher performance compared to the current state of the art techniques.

Abstract: In various implementations, a personal asset management application is configured to perform operations that facilitate the ability to search multiple images, irrespective of the images having characterizing tags associated therewith or without, based on a simple text-based query. A first search is conducted by processing a text-based query to produce a first set of result images used to further generate a visually-based query based on the first set of result images. A second search is conducted employing the visually-based query that was based on the first set of result images received in accordance with the first search conducted and based on the text-based query. The second search can generate a second set of result images, each having visual similarity to at least one of the images generated for the first set of result images.

Abstract: A data storage device includes a nonvolatile memory device including first and second memory regions; and a controller suitable for performing a pattern identification operation for write requests, based on a pattern information for one or more patterns, and storing data corresponding to the write requests in one of the first and second memory regions according to a result of the pattern identification operation.

Abstract: Some aspects are related to methods and apparatus that enable authorization of one or more functionalities of a robot based on a user physically manipulating the robot. For example, an authorization for a robot may include one or more authorization parameters associated with physical manipulation of the robot and sensor data generated in response to physical manipulation of the robot by a user may be utilized to determine if the physical manipulation conforms to the authorization parameters. If conformance is determined, one or more functionalities of the robot may be activated. Some implementations additionally and/or alternatively relate to methods and apparatus for adapting one or more physical control parameters for a robot based on preferences of a user of the robot, such as a user that is currently authorized for the robot.

Abstract: There is provided a system including a memory and a processor to receive images. The processor is further to train independent detectors for identifying first objects in the images based on individual attributes including a first attribute and a second attribute. The processor is also to train joint detectors for identifying first objects in the images based on composite attributes including a composite attributes each including the first attribute and the second attribute. The processor is to analyze features of the images to determine a difference between a first training performance of the independent detectors and a second training performance of the joint detectors. Lastly, the processor is to select, based on the analyzing, between using the independent detectors and using the joint detectors for identifying second objects in the images using a third attribute and a fourth attribute in the attribute database.