Abstract: The motion estimation of an anatomical structure may be performed using a machine-learned (ML) model trained based on medical training images of the anatomical structure and corresponding segmentation masks for the anatomical structure. During the training of the ML model, the model may be used to predict a motion field that may indicate a change between a first training image and a second training image, and to transform the first training image and a corresponding first segmentation mask based on the motion field. The parameters of the ML model may then be adjusted to maintain a correspondence between the transformed first training image and the second training image and between the transformed first segmentation mask or a second segmentation mask associated with the second training image. The correspondence may be assessed based on at least a boundary region shared by the anatomical structure and one or more other anatomical structures.

Abstract: A headphone control method and a headphone are disclosed. By simultaneously turning on an active noise cancellation link and a first transparent transmission link, external audio signals may be suppressed by generating anti-noise signals by the active noise cancellation link and playing the anti-noise signals by the speaker. Also, a non-target-type audio signal may be removed or filtered by performing noise reduction processing on the non-target-type audio signal in the external audio signals by the first transparent transmission link to obtain the noise-reduced audio signal in which the target-type audio signal in the external audio signals is retained. The speaker plays the noise-reduced audio signal, so that the target-type audio signal can be transparently transmitted to a user's ears, and the user can focus on the audio signals of the target type that they desire to perceive.

Abstract: A semiconductor device includes a first wafer, and a second wafer bonded with the first wafer and comprising a substrate and a device layer over the substrate. The first wafer is over the second wafer and includes memory arrays, array-base regions over the memory arrays, an isolation structure surrounding and insulating the array-base regions from one another, and bonding pads over the isolation structure. An array-base region of the array-base regions is connected with a memory array of the memory arrays. A bonding pad of the bonding pads is exposed.

Abstract: Embodiments of three-dimensional (3D) memory devices formed by bonded semiconductor devices and methods for forming the same are disclosed. In an example, a method for forming a semiconductor device is disclosed. The method includes the following operations. First, an insulating material layer can be formed over a substrate. In an example, single-crystalline silicon is not essential to the substrate. The insulating material layer can be patterned to form an isolation structure and a plurality of trenches in the isolation structure. A semiconductor material can be deposited to fill up the plurality of trenches to form a plurality of array-base regions in the isolation structure, the isolation structure insulating the plurality of array-base regions from one another. Further, a plurality of memory arrays can be formed over the plurality of array-base regions, and an insulating structure can be formed to cover the plurality of memory arrays and the plurality of array-base regions.

Abstract: An ionizable lipid based on endogenous dicarboxylic acid as well as a preparation method and use thereof. The ionizable lipid based on endogenous dicarboxylic acid is a compound represented by Formula (I), or a stereoisomer, tautomer, pharmaceutically acceptable salt or solvate thereof; wherein, X is selected from — Y is selected from —O— and —NH—; R is selected from C6-C20 alkyl, C6-C20 alkyl substituted by a substituent group, C6-C20 alkenyl, C6-C20 alkenyl substituted by a substituent group, C6-C20 alkynyl, and C6-C20 alkynyl substituted by a substituent group; n is an integer of 1-8.

Abstract: The motion estimation of an anatomical structure may be performed using a machine-learned (ML) model trained based on medical training images of the anatomical structure and corresponding segmentation masks for the anatomical structure. During the training of the ML model, the model may be used to predict a motion field that may indicate a change between a first training image and a second training image, and to transform the first training image and a corresponding first segmentation mask based on the motion field. The parameters of the ML model may then be adjusted to maintain a correspondence between the transformed first training image and the second training image and between the transformed first segmentation mask or a second segmentation mask associated with the second training image. The correspondence may be assessed based on at least a boundary region shared by the anatomical structure and one or more other anatomical structures.

Abstract: Embodiments of the disclosure provide a multi-class classification system. An exemplary system includes at least one processor and at least one non-transitory memory storing instructions that, when executed by the at least one processor, cause the system to perform operations. The operation includes applying a multi-class classifier to classify a set of objects into multiple classes and applying a plurality of binary classifiers to the set of objects, wherein the plurality of binary classifiers are decomposed from the multi-class classifier, each binary classifier classifying the set of the objects into a first group consisting of one or more classes selected from the multiple classes and a second group consisting of one or more remaining classes of the multiple classes. The operation also includes jointly classifying the set of objects using the multi-class classifier and the plurality of binary classifiers.

Abstract: In one embodiment, a method includes, by one or more computing systems, receiving, from a client system via an assistant xbot, a first audio input from a user, parsing the first audio input to identify a request for a content digest from an online social network, retrieving a plurality of content objects from the online social network, each of the content objects being accessible by the user, generating a customized newsfeed summary based on a determined semantical meaning of each of the plurality of content objects, and sending, to the client system via the assistant xbot, instructions for presenting an audio clip to the user responsive to the request for the content digest, wherein the audio clip comprises the customized newsfeed summary.

Abstract: The present disclosure is related to systems and methods for processing speech dialogue. The method includes obtaining target speech dialogue data. The method includes obtaining a text vector representation sequence, a phonetic symbol vector representation sequence, and a role vector representation sequence by performing a vector transformation on the target speech dialogue data based on a text embedding model, a phonetic symbol embedding model, and a role embedding model, respectively. The method includes determining a representation vector corresponding to the target speech dialogue data by inputting the text vector representation sequence, the phonetic symbol vector representation sequence, and the role vector representation sequence into a trained speech dialogue coding model. The method includes determining a summary of the target speech dialogue data by inputting the representation vector into a classification model.

Abstract: In one embodiment, a method includes receiving an indication of a trigger action by a first user at a client system, wherein the trigger action is associated with a priming content object, identifying related content objects associated with the priming content object, selecting recommended content objects based on the priming content object, the related content objects, and profile information of the first user, wherein each of the selected recommended content objects comprises entity information of entities associated with the priming content object, and presenting content suggestions at the client system, wherein each content suggestion comprises one of the selected recommended content objects.

Abstract: A self-adaptive heavy-duty gear transmission walking device includes a guide track. A carrying seat is slidably installed on the guide track. A loading platform is arranged on the carrying seat, and matching frames are arranged at the bottom of the carrying seat and positioned opposite from each other. A first driving mechanism, a walking mechanism, a first support bracket, a stop mechanism, and a second support bracket are arranged on the carrying seat. A vertical support mechanism and a horizontal support mechanism are arranged in the matching frame. The first driving mechanism is connected to one end of the walking mechanism. The walking mechanism penetrates through the carrying seat and is magnetically connected to the guide track. The other end of the walking mechanism is connected to a first transmission shaft. The first transmission shaft penetrates through the first support bracket and is fixedly installed with a brake disc.

Abstract: In some cases, lower quality, large scale training data can be automatically generated by automatic labeling. The generated training data can be used to pre-train a machine learning model. For instance, the model can be a model for detection of verbal harassment. Parameters of the pre-trained model can be refined or updated using another one or more higher-quality sets of training data, with which the model can be subsequently trained.

Abstract: In some cases, a verbal harassment detection system may use machine learning models to detect verbal harassment in real-time or near real-time. The system may receive an audio segment comprising a portion of audio captured by a microphone located within a vehicle. Further, the system may convert the audio segment to a text segment. The system may provide at least the text segment to a prediction model associated with verbal harassment detection to obtain a harassment prediction. Further, the system may provide the audio segment to an emotion detector to obtain a detected emotion of a speaking user that made an utterance included in the audio segment. Based at least in part on the harassment prediction and the detected emotion, the system may automatically, and without user intervention, determine whether a user is being harassed.

Abstract: In one embodiment, a method includes receiving an indication of a trigger action by a first user at a client system, wherein the trigger action is associated with a priming content object, identifying related content objects associated with the priming content object, selecting recommended content objects based on the priming content object, the related content objects, and profile information of the first user, wherein each of the selected recommended content objects comprises entity information of entities associated with the priming content object, and presenting content suggestions at the client system, wherein each content suggestion comprises one of the selected recommended content objects.

Abstract: In some cases, one or more heuristics can be automatically generated using a small dataset of segments previously labeled by one or more domain experts. The generated one or more heuristics along with one or more patterns can be used to assign training labels to a large unlabeled dataset of segments. A subset of segments representing an occurrence of verbal harassment can be selected using the assigned training labels. Randomly selected segments can be used as being indicative of a non-occurrence of verbal harassment. The selected subset of segments and randomly selected segments can be used to train one or more machine learning models for verbal harassment detection.

Abstract: Three-dimensional (3D) memory devices are provided. An exemplary 3D memory device includes a first semiconductor structure including a peripheral circuit, a data processing circuit, and a first bonding layer including a plurality of first bonding contacts. The peripheral circuit and the data processing circuit are stacked over one another vertically on different planes. The 3D memory device also includes a second semiconductor structure including an array of 3D NAND memory strings and a second bonding layer including a plurality of second bonding contacts. In addition, the 3D memory device includes a bonding interface between the first bonding layer and the second bonding layer. The first bonding contacts are in contact with the second bonding contacts at the bonding interface.

Abstract: A computing device receives a message including a request for a recommendation. A representation of a hypothetical ideal recommendation to provide in response to the message is determined based on the message content. Data regarding entities that are potential recommendations are retrieved from a data store, the data regarding each entity including a representation of the entity (e.g., a vector) derived from factual information about the entity and opinions of other users of the entity. Ranking scores are determined for at least a subset of the entities based on the difference between the entity representations and the representation of the hypothetical ideal recommendation. An entity to recommend is selected based on the ranking scores and a reply to the message is sent that identifies the selected entity.

Abstract: Three-dimensional (3D) memory devices are provided. An exemplary 3D memory device includes a 3D NAND memory array and an on-die data processing circuit coupled to the 3D NAND memory array on a same chip. The on-die data processing circuit is configured to receive, from an input/output (I/O) interface, control instructions for performing operations on data stored in the 3D NAND memory array. The on-die data processing circuit is also configured to retrieve the data from the 3D NAND memory array based on the control instructions and perform the operations on the retrieved data. Moreover, the on-die data processing circuit is configured to return a result of the operations to the I/O interface.

Abstract: Three-dimensional (3D) memory devices and methods for forming the same are provided. In an example, a method for forming a 3D memory device includes forming a first semiconductor structure including a peripheral circuit, a data processing circuit, and a first bonding layer including a plurality of first bonding contacts. The method also includes forming a second semiconductor structure including an array of 3D NAND memory strings and a second bonding layer including a plurality of second bonding contacts. The method further includes bonding the first semiconductor structure and the second semiconductor structure in a face-to-face manner, such that the first bonding contacts are in contact with the second bonding contacts at a bonding interface.

Abstract: The present disclosure is related to systems and methods for processing speech dialogue. The method includes obtaining target speech dialogue data. The method includes obtaining a text vector representation sequence, a phonetic symbol vector representation sequence, and a role vector representation sequence by performing a vector transformation on the target speech dialogue data based on a text embedding model, a phonetic symbol embedding model, and a role embedding model, respectively. The method includes determining a representation vector corresponding to the target speech dialogue data by inputting the text vector representation sequence, the phonetic symbol vector representation sequence, and the role vector representation sequence into a trained speech dialogue coding model. The method includes determining a summary of the target speech dialogue data by inputting the representation vector into a classification model.