Abstract: The present disclosure relates to training a machine learning model to classify images. An example method generally includes receiving a training data set including images in a first category and images in a second category. A convolutional neural network (CNN) is trained using the training data set, and a feature map is generated from layers of the CNN based on features of images in the training data set. A first area in the feature map including images in the first category and a second area in the feature map where images in the first category overlap with images in the second category are identified. The first category is split into a first subcategory corresponding to the first area and a second subcategory corresponding to the second area. The CNN is retrained based on the images in the first subcategory, images in the second subcategory, and images in the second category.

Abstract: A method, a device and a computer program product for image processing are proposed. In the method, a first training image and region information are obtained. The region information indicates a region of a defect in the first training image. A second training image with the defect at least partially removed is generated using an image generator based on the first training image and the region information. The image generator is trained to recover the first training image by replacing pixels included in the region indicated by the region information. The image generator is updated based on the second training image. In this way, the image including the defect can be accurately and efficiently recovered.

Abstract: Techniques for generating machine learning architectures are provided. A data set is received for training one or more machine learning (ML) models, where the data set comprises labeled exemplars for a plurality of classes. The data set is partitioned into a training set and a testing set. A first ML model is trained using the training set, and a quality of the first ML model with respect to each class of the plurality of classes is evaluated using the testing set. Upon determining that the quality of the first ML model is below a predefined threshold with respect to a first class and a second class of the plurality of classes, a subset of the training set is identified, where each exemplar in the subset corresponds to either the first class or the second class. A second ML model is trained using the subset of the training set.

Abstract: A method for providing a service based on a user scenario, a terminal device, and a system. The terminal device perceives the user scenario with low power consumption, and a cloud-side service open platform determines a corresponding scenario package based on the user scenario, where the scenario package includes a plurality of pre-orchestrated services for the user scenario. The service open platform pushes data related to the determined scenario package to the terminal device, and the terminal device displays entry interfaces of the plurality of services, and provides a corresponding service based on an operation performed by a user on the entry interface.

Abstract: A backlight module and a display device include a light guide plate and a point light source. The light guide plate includes a first part and a second part. The first part includes a first light incidence surface and a first light-emitting surface intersecting with the first light incidence surface. The second part includes a second light incidence surface and a second light-emitting surface intersecting with the second light incidence surface. The first light incidence surface intersects with the second light-emitting surface, the first light-emitting surface and the second light incidence surface are arranged oppositely, and both the first light-emitting surface and the second light-emitting surface include grating structures.

Abstract: Embodiments of the present disclosure relate to object defect detection. In an embodiment, a computer-implemented method is disclosed. According to the method, for a test image of at least one part of a target object, a reference image is generated by repeating a periodic pattern detected in the test image, the target object consisting of elements. A differential image is determined by comparing the test image and the reference image. The differential image is superimposed on a predefined grid image to obtain a superimposed image. The grid image comprises grids corresponding to elements of a reference object associated with the target object. The number of defective elements is determined in the at least one part of the target object based on the superimposed image. In other embodiments, a system and a computer program product are disclosed.

Abstract: A method, a device and a computer program product for image processing are proposed. In the method, whether a first image indicates a defect associated with a target object is determined. In response to determining that the first image indicates the defect, a second image absent from the defect is obtained based on the first image. The defect is identified by comparing the first image with the second image. In this way, the defect associated with the target object in the image can be accurately and efficiently identified or segmented.

Abstract: Provided is a three-dimensional (3D) display device and display method. The 3D display device includes a zoom lens unit; and a two-dimensional display unit, which is located on a side of the zoom lens unit, where a distance between the two-dimensional display unit and the zoom lens unit is less than a minimum focal length of the zoom lens unit; where the two-dimensional display unit is used for displaying multiple two-dimensional images of a three-dimensional picture, and the multiple two-dimensional images corresponds different depth of field. When the two-dimensional display unit displays the two-dimensional images corresponding to different depth of field, the zoom lens unit has a different focal length for the depth of field corresponding to each of the plurality of two-dimensional images. The display time of the three-dimensional picture is less than visual persistence time.

Abstract: Facilitating pre-positioning of streaming content onto communication devices is provided herein. A system can comprise a processor and a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations. The operations can comprise facilitating an identification of a first multimedia content stream and a second multimedia content stream. The identification can be based on a usage preference of a defined entity associated with a communication device. The operations can also comprise, in response to a determination that an available network satisfies a defined network criterion, facilitating pre-positioning, at a data store of the communication device, the first multimedia content stream and the second multimedia content stream.

Abstract: Chip package structure and chip package method are provided. The chip package structure includes an encapsulating layer, a redistribution layer, a soldering pad group, and bare chips. Connecting posts is formed on a side of the bare chips. The encapsulating layer covers the bare chips and the connecting posts, while exposes a side of the connecting posts away from the bare chips. The redistribution layer on the connecting posts includes a first redistribution wire, a second redistribution wire, and a third redistribution wire. The first redistribution wire and the second redistribution wire are electrically connected to at least one connecting post respectively, and the third redistribution layer is electrically connected to remaining connecting posts. The soldering pad group on the redistribution layer includes an input soldering pad electrically connected to the first redistribution wire and an output soldering pad electrically connected to the second redistribution wire.

Abstract: Facilitating pre-positioning of streaming content onto communication devices is provided herein. A system can comprise a processor and a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations. The operations can comprise facilitating an identification of a first multimedia content stream and a second multimedia content stream. The identification can be based on a usage preference of a defined entity associated with a communication device. The operations can also comprise, in response to a determination that an available network satisfies a defined network criterion, facilitating pre-positioning, at a data store of the communication device, the first multimedia content stream and the second multimedia content stream.

Abstract: Provided are a display module and an electronic device. The display module includes a display screen, the display screen comprising a display surface and a non-display surface opposite to the display surface; a first heat conduction member, the first heat conduction member laminated on the non-display surface, a first through hole defined in the first heat conduction member, the non-display surface comprising an exposed portion exposed in the first through hole; a photosensitive module, the photosensitive module laminated on the exposed portion; and a second heat conduction member contacting each of the photosensitive module and the first heat conduction member; the first heat conduction member comprising a first inner sidewall formed due to formation of the first through hole, and a contact area between the second heat conduction member and the photosensitive module is larger than an area of the photosensitive module directly facing the first inner sidewall.

Abstract: A driving method for an electrowetting panel is provided. The electrowetting panel includes M driving electrodes sequentially arranged along a first direction. The driving method includes providing electrical signals to the M driving electrodes, such that a droplet is acquired from a solution reservoir by the 1st driving electrode, and is driven to move by the M driving electrodes. During a droplet moving period, a pulse width of a driving signal of an mth driving electrode is Wm = ? i = 1 m ? ? W i , a pulse width of a non-driving signal between an ath driving signal and an (a+1)th driving signal of the mth driving electrode is Zma = ? i = m + 1 m + a ? ? W i . M, m, and a are positive integers, 1?m?M, and M?3. The end time of the 1st driving signal of the mth driving electrode and the end time of the mth driving signal of the 1st driving electrode are the same.

Abstract: Provided is a three-dimensional (3D) display device and display method. The 3D display device includes a zoom lens unit; and a two-dimensional display unit, which is located on a side of the zoom lens unit, where a distance between the two-dimensional display unit and the zoom lens unit is less than a minimum focal length of the zoom lens unit; where the two-dimensional display unit is used for displaying multiple two-dimensional images of a three-dimensional picture, and the multiple two-dimensional images corresponds different depth of field. When the two-dimensional display unit displays the two-dimensional images corresponding to different depth of field, the zoom lens unit has a different focal length for the depth of field corresponding to each of the plurality of two-dimensional images. The display time of the three-dimensional picture is less than visual persistence time.

Abstract: A touch-control panel includes a display panel and at least one actuator. The display panel includes a touch-control detection device. Each actuator of the at least one actuator includes a first electrode, a second electrode, and an inductive layer sandwiched between the first electrode and the second electrode. In response to the touch-control detection device detecting a touch-control operation, an actuator of the at least one actuator located in a region where the touch-control operation is performed vibrates.

Abstract: Facilitating pre-positioning of streaming content onto communication devices is provided herein. A system can comprise a processor and a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations. The operations can comprise facilitating an identification of a first multimedia content stream and a second multimedia content stream. The identification can be based on a usage preference of a defined entity associated with a communication device. The operations can also comprise, in response to a determination that an available network satisfies a defined network criterion, facilitating pre-positioning, at a data store of the communication device, the first multimedia content stream and the second multimedia content stream.

Abstract: A touch-control panel includes a display panel and at least one actuator. The display panel includes a touch-control detection device. Each actuator of the at least one actuator includes a first electrode, a second electrode, and an inductive layer sandwiched between the first electrode and the second electrode. In response to the touch-control detection device detecting a touch-control operation, an actuator of the at least one actuator located in a region where the touch-control operation is performed vibrates.

Abstract: A refrigerator includes a door having a user interface electrical module and a user interface housing fixed to a heat insulating material. The user interface housing has a first receiving space which is open towards a side of the door to receive the user interface electrical module. A handle member includes a covering portion for closing the open side of the first receiving space and a grip portion located in front of a handle groove. A rear wall of the user interface housing facing the handle groove forms a rear boundary of the handle groove. The thickness of the heat insulating material of the door which is occupied by the handle groove can therefore be reduced.

Abstract: A refrigerator includes a door having a door liner, a front wall, a user interface housing connected to the front wall, a user interface electrical module and a handle. The user interface housing, the door liner and the front wall are fixedly connected to a heat insulating layer. The user interface housing has first and second receiving spaces which can be in communication. The user interface electrical module can be inserted into the first receiving space through the second receiving space. The handle is detachably received in the second receiving space and separates the first receiving space from a handle groove in the second receiving space. The user interface housing has at least one installation wall between the first receiving space and the handle groove in a depth direction of the handle groove, and the handle is fixed to the installation wall.

Abstract: The present disclosure provides a light-emitting unit and a display device. The light-emitting includes a micro-LED element and a bonding pad. The micro-LED element includes an element pin, the bonding pad includes a bonding pin, in each welding pair of the element pin and the bonding pin, a length of the element pin is greater than a width of the bonding pin, and a non-zero angle is formed between the extending direction of the element pin and the extending direction of the bonding pin. When the element pin is deviated from the bonding pin in a certain range, the element pin can still be aligned accurately with the bonding pin. Thus, a requirement on the alignment precision between the micro-LED element and the bonding pad is lowered in a certain extent, thereby improving the yield.