Abstract: Methods and devices that include forming a first epitaxial region and a second epitaxial region above the first epitaxial region. An opening may be formed extending from the first region to the second region. And a liner layer is deposited on a sidewall and a bottom of the opening. A plasma treatment is performed on the liner layer, which can form a conditioned or passivated region of the first epitaxial region that may be maintained during the growth of additional epitaxial material on the second epitaxial region.

Abstract: The present disclosure discloses an adaptive high-precision compression method and system based on a convolutional neural network model, and belongs to the fields of artificial intelligence, computer vision, and image processing. According to the method of the present disclosure, coarse-grained pruning is performed on a neural network model by using a differential evolution algorithm first, and the coarse-grained space is quickly searched through an entropy importance criterion and an objective function with good guidance to obtain a near-optimal neural network structure. Then fine-grained search space is built on the basis of an optimal individual obtained from the coarse-grained search, and fine-grained pruning is performed on the neural network model by a differential evolution algorithm to obtain a network model with an optimal structure. Finally, the performance of the optimal model is restored by using a multi-teacher multi-step knowledge distillation network to reach the precision of an original model.

Abstract: A method includes forming a device layer on a first surface of a first substrate, forming a first interconnect structure over the device layer, depositing a bonding layer over the first interconnect structure, forming a diamond layer over the bonding layer, performing a laser treatment on a top portion of the diamond layer by applying laser energy to the top portion of the diamond layer using a laser beam, and performing a thinning process on the diamond layer to remove the top portion of the diamond layer.

Abstract: In an embodiment, a method includes forming an opening in a dielectric layer; filling the opening with a conductive material; and performing a chemical mechanical polishing process on the conductive material and the dielectric layer, the chemical mechanical polishing process comprising a slurry, the slurry comprising: abrasives, the abrasives comprising titania-silica hybrid particles; and an oxidizer.

Abstract: A method includes performing a cleaning process on a first surface of a first wafer, and performing a surface activation process on the first surface. The surface activation process is selected from the group consisting of: a plasma surface activation process comprising generating a plasma from a process gas, wherein ions in the plasma are removed using a filter, and wherein a remaining uncharged part of the plasma is used to treat the first surface; a laser surface activation process using a laser beam; an acid surface activation process using an acid; and an alkali surface activation process using an alkali. After the surface activation process, a rinsing process is performed on the first surface. The first surface of the first wafer is bonded to a second surface of a second wafer.

Abstract: A semiconductor apparatus for pre-wetting a semiconductor workpiece includes a process chamber, a workpiece holder disposed within the process chamber to hold the semiconductor workpiece, a pre-wetting fluid tank disposed outside the process chamber and containing a pre-wetting fluid, and a conduit coupled to the pre-wetting fluid tank and extending into the process chamber. The conduit delivers the pre-wetting fluid from the pre-wetting fluid tank out through an outlet of the conduit to wet a major surface of the semiconductor workpiece, wherein the outlet of the conduit is positioned above the major surface of the semiconductor workpiece by a vertical distance.

Abstract: A method includes forming a lower semiconductor region, forming an upper semiconductor region overlapping the lower semiconductor region, forming a lower gate dielectric and an upper gate dielectric on the lower semiconductor region and the upper semiconductor region, respectively, forming a lower gate electrode on the lower gate dielectric and the upper gate dielectric, etching back the lower gate electrode, forming a gate isolation layer on the lower gate electrode that has been etched back, and forming an upper gate electrode over the gate isolation layer. The upper gate electrode is on the upper gate dielectric.

Abstract: A master node (MN) that determines a manner of a protocol data unit (PDU) session to be served, wherein the manner of the PDU session includes a split PDU session between the MN and a secondary node (SN) or a whole PDU session served by the MN, generates a request message to be transmitted to the SN to setup PDU session resources on the SN based on the manner of the PDU session is to be served and indicates, to the SN, the manner of the PDU session to be served. A SN that receives, from a MN, a request message to setup PDU session resources, determines a manner of the PDU session to be served, and sets up the PDU session resources on the SN, based on at least the manner of the PDU session to be served.

Abstract: A method includes putting a first package component into contact with a second package component. The first package component comprises a first dielectric layer including a first dielectric material, and the first dielectric material is a silicon-oxide-based dielectric material. The second package component includes a second dielectric layer including a second dielectric material different from the first dielectric material. The second dielectric material comprises silicon and an element selected from the group consisting of carbon, nitrogen, and combinations thereof. An annealing process is performed to bond the first dielectric layer to the second dielectric layer.

Abstract: High resolution 3D thermal imaging can be obtained by using enhanced non-destructive heat transducer designs. A thermal property measurement method includes providing a sample for thermal property measurement, and bonding a transducer layer on the sample through a temporary bonding layer. Thermal measurement processes are performed along the X-Y, X-Z and Y-Z planes of the sample, wherein the X-Y plane is parallel to a top surface of the sample, and the X-Z plane and Y-Z plane are perpendicular to the top surface of the sample. Each thermal measurement processes include heating a designated region of the sample covered with the transducer layer using a pump laser, and using a probe laser for generating a reflectance signal of the sample, and determining a thermal conductivity in the designated region of the sample from the reflectance signal. Furthermore, the transducer layer is removed along with the temporary bonding layer from the sample.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for fine-tuning diffusion-based generative neural networks in compact parameter spaces for text-to-image generation. In one aspect, a method performed by one or more computers for fine-tuning a diffusion-based generative neural network to obtain a fine-tuned version of the diffusion-based generative neural network is described.

Abstract: Methods, systems, and computer programs encoded on a computer storage medium, that relate to extracting digital watermarks from images, irrespective of distortions introduced into these images. Methods can include inputting a first data item into a channel encoder that can generate a first encoded data item that is greater in length than the first data item and that (1) includes the input data item and (2) new data this is redundant of the input data item. Based on the first encoded data item and a first image, an encoder model can generate a first encoded image into which the first encoded data is embedded as a digital watermark. A decoder model can decode the first encoded data item to generate a second data, which can be decoded by the channel decoder to generate data that is predicted to be the first data.

Abstract: Provided is an air treatment device (1000). The air treatment device (1000) includes a housing (50) having a first air duct configured to supply air and a second air duct configured to discharge air and a rotary wheel assembly (100) rotatably disposed in the housing (50). The rotary wheel assembly (100) has an adsorption member (30) capable of adsorbing and releasing moisture. The rotary wheel assembly (100) is obliquely disposed in the housing (50).

Abstract: Techniques for determining perceptual quality indicators of video content items are provided. In some embodiments, a system including one or more processors executes instructions to: receive a video content item comprising a plurality of frames; determine, using a first subnetwork of a deep neural network, a content quality indicator for each frame, wherein the content quality indicator corresponds to one or more semantic content indicators for each frame; determine, using a second subnetwork of the deep neural network, a video distortion indicator for each frame, wherein the video distortion indicator indicates a quality of the frame based on distortions contained within the frame; generate a quality level for each frame based on at least the content quality indicator and the video distortion indicator for the frame; and output an indication of quality for the video content item based on the quality level for each frame.

Abstract: A monocular video-based three-dimensional reconstruction and depth prediction method and system for a pipeline are provided, relating to the technical field of deep learning and image processing. The method includes the following steps: performing feature extraction and matching and incremental reconstruction on collected image sequences of different pipeline scenes based on the COLMAP method to obtain a camera external parameter matrix corresponding to each image, and constructing a pipeline three-dimensional reconstruction dataset; training the Fast-MVSNet network and the PatchMatchNet network using the existing public datasets, and selecting an optimal network model from multiple trained network models for three-dimensional reconstruction and depth prediction of an actual pipeline scene.

Abstract: Systems, devices, methods, and computer readable medium for evaluating visual quality of digital content are disclosed. Methods can include training machine learning models on images. A request is received to evaluate quality of an image included in a current version of a digital component generated by the computing device. The machine learning models are deployed on the image to generate a score for each quality characteristic of the image. A weight is assigned to each score to generate weighted scores. The weighted scores are combined to generate a combined score for the image. The combined score is compared to one or more thresholds to generate a quality of the image.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for detecting and decoding a visually imperceptible or perceptible watermark. A watermark detection apparatus determines whether the particular image includes a visually imperceptible or perceptible watermark using detector a machine learning model. If the watermark detection apparatus detects a watermark, the particular image is routed to a watermark decoder. If the watermark detection apparatus cannot detect a watermark in the particular image, the particular image is filtered from further processing. The watermark decoder decodes the visually imperceptible or perceptible watermark detected in the particular image. After decoding, an item depicted in the particular image is validated based data extracted from the decoded visually imperceptible or perceptible watermark.

Abstract: Disclosed are an SRv6 message processing method and apparatus, a communication device, and a storage medium. The method comprises: a user side equipment setting a source address in an outer layer message header in a first SRv6 message as a first segment identifier (SID) on the basis of a pre-obtained first SID used for representing an address, and generating a second SRv6 message, where the first SID is pre-planned by a network service provider; and sending the second SRv6 message to a network device.

Abstract: A method includes forming a multi-layer stack including a plurality of semiconductor nanostructures. The multi-layer stack includes a semiconductor nanostructure, and a sacrificial semiconductor layer over the semiconductor nanostructure. The method further includes depositing a semiconductor layer over and contacting the semiconductor nanostructure, removing the sacrificial semiconductor layer, and forming a replacement gate stack encircling a combined region of the semiconductor nanostructure and the semiconductor layer.

Abstract: A computer-implemented method that provides watermark-based image reconstruction to compensate for lossy encoding schemes. The method can generate a difference image describing the data loss associated with encoding an image using a lossy encoding scheme. The difference image can be encoded as a message and embedded in the encoded image using a watermark and later extracted from the encoded image. The difference image can be added to the encoded image to reconstruct the original image. As an example, an input image encoded using a lossy JPEG compression scheme can be embedded with the lost data and later reconstructed, using the embedded data, to a fidelity level that is identical or substantially similar to the original.