Abstract: A method for classifying activity based on multi-sensor input includes receiving, from two or more sensors, sensor data indicating activity within a building, determining, for each of the two or more sensors and based on the received sensor data, (i) an extracted feature vector for activity within the building and (ii) location data, labelling each of the extracted feature vectors with the location data, generating, using the extracted feature vectors, an integrated feature vector, detecting a particular activity based on the integrated feature vector, and in response to detecting the particular activity, performing a monitoring action.

Abstract: The present disclosure provides methods and an apparatuses for building a data identification model. One exemplary method for building a data identification model includes: performing logistic regression training using training samples to obtain a first model, the training samples comprising positive and negative samples; sampling the training samples proportionally to obtain a first training sample set; identifying the positive samples using the first model, and selecting a second training sample set from positive samples that have identification results after being identified using the first model; and performing Deep Neural Networks (DNN) training using the first training sample set and the second training sample set to obtain a final data identification model. The methods and the apparatuses of the present disclosure improve the stability of data identification models.

Abstract: A mobile terminal is a communication device that communicates using visible light, and includes: a plurality of light emitters, each of which is disposed along the front surface of the mobile terminal and transmits a visible light signal; and a plurality of light receivers, each of which is disposed along the front surface and receives a visible light signal. The plurality of light emitters are disposed on the left side of all of the plurality of light receivers, and the plurality of first light receivers are disposed on the right side of all of the plurality of light emitters.

Abstract: Providing a task-aware recommendation of hyperparameter configurations for a neural network architecture. First, a joint space of tasks and hyperparameter configurations are constructed using a plurality of tasks (each of which corresponds to a dataset) and a plurality of hyperparameter configurations. The joint space is used as training data to train and optimize a performance prediction network, such that for a given unseen task corresponding to one of the plurality of tasks and a given hyperparameter configuration corresponding to one of the plurality of hyperparameter configurations, the performance prediction network is configured to predict performance that is to be achieved for the unseen task using the hyperparameter configuration.

Abstract: A method may include training a machine learning model to perform a first task before training the machine learning model to perform the second task. The machine learning model includes a generator network and a discriminator network. The training includes training, based on a first training sample associated with the first task, the discriminator network to perform the first task. The generator network may be trained to generate a first synthetic training sample emulating the first training sample. The discriminator network trained to perform the first task may be reinitialized in order for the discriminator network to be trained, based on a second training sample, to perform the second task. The reinitialized discriminator network may be further retrained, based on the first synthetic training sample, to perform the first task. Related systems and articles of manufacture, including computer program products, are also provided.

Abstract: Examples described herein relate to automatic identification and transformation of a color region. A user can identify a region of a video frame or image that corresponds to a color region that is to be segmented. A color region can include one or more colors that appear to be approximately a uniform color. For one or more video frames, gamma correction can be applied to frames of the video. One or more frames of a video can be mapped to two color spaces. For each pixel in an image, a determination is made if the pixel has the same color as that of the identified region based on each of the at least two color spaces identifying the pixel as the color. The color region can be identified throughout a video and transformed to another color to aid in video editing.

Abstract: Enhancing color reproduction of an image in an upscaling process, the method comprising: converting RGB-formatted data into luminance channel formatted data and color channel formatted data; converting RGB-predicted data into luminance channel predicted data and a color channel predicted data; computing a first loss between the RGB-formatted data and the RGB-predicted data; computing a second loss between the color channel formatted data and the color channel predicted data; and outputting a weighted average value of the first loss and the second loss to enhance the color reproduction of the image.

Abstract: Concepts and technologies directed to optical networking with hybrid optical vortices are disclosed herein. Embodiments can include a system that is configured to perform operations for optical networking with hybrid optical vortices. The system can include a hybrid optical switch that can communicatively couple with another network device via one or more nanofiber communication paths. The operations can include receiving, from a first nanofiber communication path, a hybrid optical vortex that carries an internet protocol packet. The operations also can include decoupling the hybrid optical vortex to extract an optical vortex that encapsulates the internet protocol packet. The operations also can include switching the internet protocol packet to a subsequent communication path based on the optical vortex that encapsulates the internet protocol packet.

Abstract: A landslide recognition method based on Laplacian pyramid remote sensing image fusion includes: performing original remote sensing image reconstruction based on extracted local features and global features of remote sensing images through a Laplacian pyramid fusion module to generate a fused image, constructing a deep learning semantic segmentation model through a semantic segmentation network, labeling the fused image to obtain a dataset of landslide disaster label map, and training the deep learning semantic segmentation model by the dataset, and then storing when a loss curve is fitted and a landslide recognition accuracy of remote sensing image of the deep learning semantics segmentation model meets a requirement by modifying a structure of the semantic segmentation network and adjusting parameters of the deep learning semantics segmentation model.

Abstract: A safety-based prediction apparatus, system and method are provided. A machine learning hardware accelerator (MLHA) includes a main classifier (MC) module, at least one guardian classifier (GC) module, and a final predicted class decision module. The MC module predicts an MC predicted class based on input data, and includes a pre-trained, machine learning main classifier (MLMC) that has at least one safety critical (SC) class and a plurality of non-SC classes. Each guardian classifier (GC) module is associated with an SC class, and predicts a GC predicted class based on the input data. Each GC module includes a pre-trained, machine learning guardian classifier (MLGC) having two classes including an associated SC class and a residual class that includes any non-associated SC classes and the plurality of non-SC classes. A decision module determines and outputs a final predicted class based on the MC predicted class and each GC predicted class.

Abstract: Methods, systems, and computer-readable storage media for tuning behavior of a machine learning (ML) model by providing an alternative loss function used during training of a ML model, the alternative loss function enhancing reliability of the ML model, calibrating the confidence of the ML model after training, and reducing risk in downstream tasks by providing a mapping between the confidence of the ML model to the expected accuracy of the ML model.

Abstract: A communication system transmits relay data through a communication path including a plurality of sections in which different communication schemes are used. A relay communication device is provided between a first section and a second section which are adjacent sections. The relay communication device includes a receiving unit receiving the relay data from the first section through a frame of a first communication scheme, and a relaying unit configuring, in a frame of a second communication scheme used to transmit the relay data to a relay destination, signal quality information representing signal quality calculated for a physical link in each of the sections through which the relay data is transmitted before arriving at the relay communication device, and outputting the frame of the second communication scheme to the second section.

Abstract: An optical wireless transmission system 10 includes a transmission device including at least one memory storing instructions, and at least one processor configured to execute the instructions to; generate a plurality of digital outphasing signals; orthogonally modulate the digital outphasing signals at an intermediate frequency; and set an intermediate frequency for satisfying a specified signal-to-distortion power ratio based on a sampling frequency, wherein the digital outphasing signals are orthogonally modulated at the intermediate frequency; a hardware optical fiber module configured to convert orthogonally modulated digital electrical signals into optical signals, transmit the optical signals through an optical fiber, and convert the optical signals into digital electrical signals; and a remote unit configured to combine the digital electrical signals transmitted by the hardware optical fiber module, and transmit a combined signal as a radio signal.

Abstract: An image processing apparatus decodes encoded RAW data that includes subband data being encoded with lossy encoding scheme, and determines one of a plurality of classifications based on the decoded subband data, wherein the plurality of classifications are based on a feature of an image. The apparatus also obtains correction data corresponding to the determined classification, and corrects recomposed data, which is obtained by applying frequency recomposition to the decoded subband data, based on the correction data, in order to obtain the corrected data as decoded RAW data.

Abstract: An image processing apparatus comprising a first correction unit which performs first gray-level correction on an image of a target region using a first correction amount, a first estimation unit which estimates a number of specific objects in the image of the target region corrected by the first correction unit, a second correction unit which performs second gray-level correction on the image of the target region using a second correction amount that is different from the first correction amount, a second estimation unit which estimates the number of specific objects in the image of the target region corrected by the second correction unit; and a selection unit which selects a larger one of the number of the specific objects estimated by the first estimation unit and the number of the specific objects estimated by the second estimation unit.

Abstract: At least two cascaded chains of intelligent support boxes linked by optical cable provide the needed elements to link the electric AC power line and low voltage DC power lines to the many known outlets and switches (wiring devices) and their loads be it AC or DC operated, with the AC and DC lines are separately drawn through separated conduits and ducts safely and accurately controlled by an home controller and/or via a command converter and a distributor.

Abstract: A laser device for optical communication comprises a first laser unit connected to a first optical fiber for supplying a transmission laser beam thereto. wherein the laser device is configured for providing a reference laser beam in addition to the transmission laser beam. For providing the reference laser beam the laser device further includes a second laser unit connected to a second optical fiber for supplying the reference laser beam to the second optical fiber. The first laser unit is configured for providing the transmission laser beam as a linear polarized beam that is polarized in a first polarization direction, and the second laser unit is configured for providing the reference laser beam as a linear polarized beam that is polarized in a second polarization direction.

Abstract: A data compression device and a compression method are provided. The data compression device includes a quantization table processing unit and a quantization unit. The quantization table processing unit determines a target quantization table in which a quantization coefficient satisfies a data distortion rate and a compression ratio of a preset condition according to a target compression ratio. By constructing different quantization tables for different data, a distortion rate is greatly reduced based on satisfying a compression ratio, and issues that the distortion rate and the compression ratio cannot be simultaneously satisfied in the prior art are alleviated.

Abstract: Techniques for coding sets of images with neural networks include transforming a first image of a set of images into coefficients with an encoder neural network, encoding a group of the coefficients as an integer patch index into coding table of table entries each having vectors of coefficients, and storing a collection of patch indices as a first coded image. The encoder neural network may be configured with encoder weights determined by jointly with corresponding decoder weights of a decoder neural network on the set of images.

Abstract: An image processing apparatus includes one or more processors including hardware. The one or more processors are configured to: divide, regarding a plurality of images acquired by capturing images of an imaging subject over time, the respective images into a plurality of regions; calculate, for the respective divided regions, motion vectors by detecting movements of the imaging subject; detect luminances and contrasts of the respective regions and detect specific regions where the detected luminances and contrasts satisfy determination conditions, the determination conditions being that the luminances are equal to or greater than a first threshold and that the contrasts are equal to or less than a second threshold; select the motion vectors to be used by excluding the motion vectors of the detected specific regions from the calculated motion vectors; and generate a combined image by performing position alignment of the plurality of images by using the selected motion vectors.