Abstract: A memory expansion device includes an expansion control circuit configured to receive a near data processing (NDP) request and a remote memory device configured to store data corresponding to the NDP request according to control by the expansion control circuit. In response to the NDP request, the expansion control circuit performs a request processing operation to perform a read or a write operation on the remote memory device and performs a computation operation using the data corresponding to the NDP request.

Abstract: A semiconductor device includes a substrate, a bit line extending in a first direction on the substrate, a first vertical channel pattern and a second vertical channel pattern on the bit line, a back gate electrode between the first vertical channel pattern and the second vertical channel pattern and extending in a second direction perpendicular to the first direction across the bit line, a first word line extending in the second direction from one side of the first vertical channel pattern, a second word line extending in the second direction from other side of the second vertical channel pattern, and a contact pattern connected to each of the first vertical channel pattern and the second vertical channel pattern. When viewed from a cross-sectional view, each of the first vertical channel pattern and the second vertical channel pattern have a trapezoidal shape with the long sides facing each other.

Abstract: A system stores first and second images generated by first and second cameras; applies a segmentation model to the first image to generate a first segmentation mask identifying object instances; applies the segmentation model to the second image to generate a second segmentation mask identifying the object instances; projects the first segmentation mask to a viewpoint of the second camera to generate a first projected segmentation mask; converts the first projected segmentation mask and the second segmentation mask to first and second semantic masks, respectively; and computes a first similarity value based on the first and second semantic masks. This may be repeated exchanging the first and second images to compute a second similarity value. The system determines a loss value based on the first similarity value and the second similarity value and trains the segmentation model based on the loss value.

Abstract: A semiconductor package includes: a first semiconductor chip; a second semiconductor chip stacked on the first semiconductor chip; an underfill material layer interposed between the first semiconductor chip and the second semiconductor chip; and a first dam structure disposed on the first semiconductor chip. The first dam structure extends along an edge of the second semiconductor chip and includes unit dam structures apart from each other with a slit therebetween. A vertical level of an upper surface of the first dam structure is located between a vertical level of a lower surface of the second semiconductor chip and a vertical level of an upper surface of the second semiconductor chip. A first sidewall of the first dam structure is in contact with the underfill material layer and includes a flat surface parallel to a sidewall of the second semiconductor chip that faces the first sidewall of the first dam structure.

Abstract: An example system include one or more memories and one or more processors coupled to the one or more memories. The one or more processors are configured to perform a first segmentation operation on a previous frame of image data to generate first segmentation data. The one or more processors are configured to perform a deformable convolution operation based on the first segmentation data to generate a deformable convolution output. The one or more processors are configured to perform a second segmentation operation on a current frame of image data to generate second segmentation data. The one or more processors are configured to combine the deformable convolution output with the second segmentation data to generate third segmentation data and control operation of a device based on the third segmentation data.

Abstract: There is provided a storage device, which includes: a memory device that includes a plurality of memory blocks, and stores first meta data including first status data and a first parameter in a first memory block among the plurality of memory blocks; and a memory controller that stores second meta data including second status data and second parameters, determines final meta data among a plurality of pieces of meta data including the first meta data and the second status data by comparing a plurality of pieces of status data with the first status data and the second status data, performs parameter confirmation for storing the final meta data in the meta block, and controls the memory device based on a parameter stored in the meta block.

Abstract: Radio frequency (RF) sensing of a target object by a wireless network is supported by a sensing node and a network node using sensing assistance data. The sensing assistance data may include information for Doppler estimation from received RF signals. The information for Doppler estimation, for example, may include an indication of phase coherency of RF signals, such as phase coherence found in positioning reference signal (PRS) resources, resource sets, transmission reception points (TRPs), frequency layer or any combination thereof. The sensing assistance data may include an association of PRS resources and reference point objects (RPOs) so RF signals reflected from the RPOs may be excluded from the sensing measurements. The sensing assistance data may include beam pattern information for PRS resources for beam coordination for RF sensing.

Abstract: The present disclosure relates to a communication method and system for converging a 5th-Generation (5G) communication system or a 6th-Generation (6G) communication system for supporting higher data rates beyond a 4th-Generation (4G) system. Provided is a method, system, and apparatus for sharing a dynamic frequency in a wireless communication system. A method, performed by a network entity, of assigning a frequency channel, includes: receiving a frequency channel assignment request from a first device; identifying a device cluster including the first device; and assigning, to the first device, at least one frequency channel included in a frequency channel set assigned to the device cluster including the first device. The device cluster may satisfy a certain condition and may be a set of a plurality of devices sharing at least one frequency channel included in the frequency channel set.

Abstract: A semiconductor memory device includes: a plurality of word lines extending in a first direction; a plurality of channel layers alternately arranged with the word lines, wherein the channel layers extend in a second direction; and a plurality of bit lines located on the word lines and the channel layers, and extending in a third direction. The bit lines are electrically connected to the channel layers. The plurality of word lines include: a plurality of selected word lines to which a positive voltage is applied; and a plurality of non-selected word lines to which a negative voltage is applied. The plurality of channel layers include: a selected channel layer that is turned on; and a non-selected channel layer that is turned off.

Abstract: A method for providing content using a marker by an electronic device is provided. The method comprise obtaining an image for a first marker through a camera, obtaining a first content corresponding to the first marker and a position of the first marker, based on the image, obtaining a second content corresponding to a second marker and a position of the second marker, based on information received from an external electronic device, and obtaining a third content by combining the first content and the second content based on the position of the first marker and the position of the second marker.

Abstract: Disclosed is a method performed by a terminal in a wireless communication system, including performing rate matching for a code block based on a circular buffer, and transmitting, to a base station, uplink data based on the rate matching, wherein, in case that the rate matching is limited buffer rate matching (LBRM), a length of the circular buffer is determined based on a value obtained by dividing a transport block size (TBS) associated with the LBRM by a product of a number of code blocks of a transport block and a code rate associated with the LBRM.

Abstract: A semiconductor memory device includes a memory cell array, an error correction code (ECC) circuit, a fault address register and a control logic circuit. The memory cell array includes a plurality of memory cell rows. The scrubbing control circuit generates scrubbing addresses for performing a scrubbing operation on a first memory cell row based on refresh row addresses for refreshing the memory cell rows. The control logic circuit controls the ECC circuit such that the ECC circuit performs an error detection and correction operation on a plurality of sub-pages in the first memory cell row to count a number of error occurrences during a first interval and determines a sub operation in a second interval in the scrubbing operation based on the number of error occurrences in the first memory cell row.

Abstract: A gas diffusion layer for anion exchange membrane electrolysis which is hydrophilic and exhibits gas permeability is disclosed. A method for manufacturing the gas diffusion layer for anion exchange membrane electrolysis is disclosed. The gas diffusion layer for anion exchange membrane electrolysis contains a reticular carbon nanofiber film, wherein the carbon nanofiber has a diameter of 500 nm or less.

Abstract: Certain aspects of the present disclosure provide techniques for training and using an instance segmentation neural network to detect instances of a target object in an image. An example method generally includes generating, through an instance segmentation neural network, a first mask output from a first mask generation branch of the network. The method further includes generating, through the instance segmentation neural network, a second mask output from a second, parallel, mask generation branch of the network. The second mask output is typically of a lower resolution than the first mask output. The method further includes combining the first mask output and second mask output to generate a combined mask output. Based on the combined mask output, an output of the instance segmentation neural network is generated. One or more actions are taken based on the generated output.

Abstract: An inductively coupled plasma processing apparatus includes a main body having an internal space; a chamber in the internal space of the main body and having an open upper side and an open lower side; a window unit coupled to an upper portion of the chamber to form a processing space; and a coil in an upper portion of the window unit, wherein the coil is configured to form an electromagnetic field. The main body includes a substrate installation portion configured to receive a substrate so that the substrate is below the chamber. The window unit includes a plurality of windows, and each of the plurality of windows has a respective different thickness and/or a respective different material.

Abstract: Techniques and systems are provided for generating one or more segmentations masks. For instance, a process may include generating a delta image based on a difference between a current image and a prior image. The process may further include processing, using a transform operation, the delta image and features representing the prior image to generate a transformed feature representation of the prior image. The process may include combining the transformed feature representation of the prior image with features representing the current image to generate a combined feature representation of the current image. The process may further include generating, based on the combined feature representation of the current image, a segmentation mask for the current image.

Abstract: Certain aspects of the present disclosure provide techniques and apparatus for regenerative learning to enhance dense predictions. In one example method, an input image is accessed. A dense prediction output is generated based on the input image using a dense prediction machine learning (ML) model, and a regenerated version of the input image is generated. A first loss is generated based on the input image and a corresponding ground truth dense prediction, and a second loss is generated based on the regenerated version of the input image. One or more parameters of the dense prediction ML model are updated based on the first and second losses.

Abstract: Techniques and systems are provided for performing online adaptation of machine learning model(s). For example, a process may include obtaining features extracted from a image by a machine learning model during inference and determining, by the machine learning model based on the features during inference, a plurality of keypoint estimates in the image and/or a bounding region estimate associated with an object in the image. The process may further include generating pseudo-label(s) based on the plurality of keypoint estimates and/or the bounding region estimate. The process may include determining at least one self-supervised loss based on the plurality of keypoint estimates and/or the bounding region estimate. The process may further include adapting, based on the at least one self-supervised loss, parameter(s) of the machine learning model. The process may include generating, using the machine learning model with the adapted parameter(s), a segmentation mask for the image (or another image).

Abstract: Systems and techniques are provided for processing data (e.g., image data). For instance, according to some aspects of the disclosure, a method may include receiving, at a transformer of a machine learning system, learnable queries, keys, and values obtained from a feature map of a segmentation model of the machine learning system. The method may further include learning, via the transformer, a mapping between an unsupervised output and a supervised output of the segmentation model based on the feature map.