Abstract: This application provides a face living body detection method performed by a computing device, the method including: obtaining a first face image of a target detection object in a first illumination condition and a second face image of the target detection object in a second illumination condition, determining a difference image according to the two images, decoupling an object reflectivity and an object normal vector corresponding to the target detection object from a feature map extracted from the difference image, and determining whether the target detection object is a living body according to the object reflectivity and the object normal vector. This method decouples texture information and depth information of a face, and performs living body detection by using decoupled information, which increases the defense capability against 3D attacks, thereby effectively defending against planar attacks and 3D attacks.

Abstract: A method of training a machine learning algorithm comprises providing a set of input data, performing transforms on the input data to generate augmented data, to provide transformed base paths into machine learning algorithm encoders, segmenting the augmented data, calculating main base path outputs by applying a weighting to the segmented augmented data, calculating pruning masks from the input and augmented data to apply to the base paths of the machine learning algorithm encoders, the pruning masks having a binary value for each segment in the segmented augmented data, calculating sparse conditional path outputs by performing a computation on the segments of the segmented augmented data, and calculating a final output as a sum of the main base path outputs and the sparse conditional path outputs. A computer-implemented system for learning sparse features of a dataset is also disclosed.

Abstract: A method for increasing the temperature-resiliency of a neural network, the method comprising loading a neural network model into a resistive nonvolatile in-memory-computing chip, training the deep neural network model using a progressive knowledge distillation algorithm as a function of a teacher model, the algorithm comprising injecting, using a clean model as the teacher model, low-temperature noise values into a student model and changing, now using the student model as the teacher model, the low-temperature noises to high-temperature noises, and training the deep neural network model using a batch normalization adaptation algorithm, wherein the batch normalization adaptation algorithm includes training a plurality of batch normalization parameters with respect to a plurality of thermal variations.

Abstract: A spectrometer and a fabrication method thereof. The spectrometer includes: a first base substrate; a second base substrate opposite to the first base substrate; a detection channel between the first base substrate and the second base substrate; a quantum dot light emitting layer on a side of the first base substrate that is close to the second base substrate, and including a plurality of quantum dot light emitting units; a black matrix on the side of the first base substrate that is close to the second base substrate, and configured to separate the plurality of quantum dot light emitting units; and a sensor layer, including a plurality of sensors, the plurality of sensors being in one-to-one correspondence with the plurality of quantum dot light emitting units.

Abstract: Disclosed is an apparatus for charging a battery comprising a first charging device configured to communicate with at least one second charging device, the first charging device and the at least one second charging device configured to charge the battery, and comprising a first controller configured to control the first charging device, wherein the first controller determines the number of the at least one second charging devices by communicating with a second controller.

Abstract: A dynamic modular neural network (DMNN) for NOx emission prediction in MSWI process is provided. First, the input variables are smoothed and normalized. Then, a feature extraction method based on principal component analysis (PCA) was designed to realize the dynamic division of complex conditions, and the prediction task to be processed was decomposed into sub-tasks under different conditions. In addition, aiming each sub-tasks, a long short-term memory (LSTM)-based sub-network is constructed to achieve accurate prediction of NOx emissions under various working conditions. Finally, a cooperative strategy is used to integrate the output of the sub-networks, further improving the accuracy of prediction model. Finally, merits of the proposed DMNN are confirmed on a benchmark and real industrial data of a municipal solid waste incineration (MSWI) process. The problem that the NOx emission of MSWI process is difficult to be accurately predicted due to the sensor limitation is effectively solved.

Abstract: A system for repairing a valve leaflet in minimally invasive surgery includes: a delivery device comprising: an operating handle; an outer delivery tube coupled to the operating handle; a delivery needle coupled to the operating handle; and a pushing tube coupled to the operating handle. The system further comprises a repair component comprising an anchor and a connection wire. The delivery device is configured to perform, in response to a trigger operation, a linkage release operation that includes actuating the delivery needle and releasing the repair component.

Abstract: Disclosed is an atrial appendage clip, including a first clip arm and a second clip arm that are symmetrically arranged; a first cavity and a second cavity are correspondingly provided in the first clip arm and the second clip arm along an axial direction thereof; a first spring and a second spring are correspondingly provided in the first cavity and the second cavity; the first spring is connected with the second clip arm via a first pull wire, and the second spring is connected with the first clip arm via a second pull wire, so as to form a closed loop structure; and as the first spring and the second spring are stretched or compressed, the first clip arm and the second clip arm are opened or closed therebetween.

Abstract: Hardware noise-aware training for improving accuracy of in-memory computing (IMC)-based deep neural network (DNN) hardware is provided. DNNs have been very successful in large-scale recognition tasks, but they exhibit large computation and memory requirements. To address the memory bottleneck of digital DNN hardware accelerators, IMC designs have been presented to perform analog DNN computations inside the memory. Recent IMC designs have demonstrated high energy-efficiency, but this is achieved by trading off the noise margin, which can degrade the DNN inference accuracy. The present disclosure proposes hardware noise-aware DNN training to largely improve the DNN inference accuracy of IMC hardware. During DNN training, embodiments perform noise injection at the partial sum level, which matches with the crossbar structure of IMC hardware, and the injected noise data is directly based on measurements of actual IMC prototype chips.

Abstract: Aspects of the subject disclosure may include, for example, a device comprising: a first micro-camera-element comprising a first sensor area and a first aperture element, the first aperture element having a first structural configuration, the first aperture element and the first sensor area being disposed relative to each other in order to cooperate in obtaining first imaging data having first characteristics, and the first characteristics comprising first imaging resolution and first angular coverage; a second micro-camera-element comprising a second sensor area and a second aperture element, the second aperture element having a second structural configuration, the second aperture element and the second sensor area being disposed relative to each other in order to cooperate in obtaining second imaging data having second characteristics, the second characteristics comprising second imaging resolution and second angular coverage, and the first imaging resolution differing from the second imaging resolution, th

Abstract: A knotter, including: a driving structure; a punching and shearing clip, a clip mouth of the punching and shearing clip having a first position in an open state, and a second position in a punching state when driven by the driving structure; and scissor having a first position in an open state when driven by the driving structure, and a second position in a shearing state when driven by the driving structure. The second position of the scissor is synchronized with the second position of the punching and shearing clip. The disclosure effectively solves the problem that the shearing knife of the knotter in the related art is unstable when cutting the suture line.

Abstract: An atrial appendage occluder capable of entering a half-released state by means of pushing a head-end fiber comprises: an occlude body (1); a head-end control fiber (2); and a tail-end control fiber (3). A head end of the occluder body (1) is connected to one end of the head-end control fiber (2) by means of a head-end threaded bushing (4). A tail end of the occluder body (1) is connected to one end of the tail-end control fiber (3) by means of a tail-end threaded bushing (5). The tail-end control fiber (3) is in the form of a hollow column. The other end of the head-end control fiber (2) sequentially passes through the tail-end threaded bushing (5) and the tail-end control fiber (3). The occluder body (1) is in a woven-net support structure, and has a shape preconfigured to match the structure of an atrial appendage after the occluder body (1) has been fully released. The occluder body (1) is in the form of a strip and disposed in an outer sheath (6) before being released.

Abstract: Disclosed is an atrial appendage clip delivery system, which includes a handle shell, a push rod, a connection pipe and a support base; a passage is provided in the handle shell along a length direction; one end of the push rod is slidably disposed in the passage; the support base is provided with a mounting window for mounting an atrial appendage clip; the atrial appendage clip includes a first clip arm and a second clip arm capable of opening in parallel or closing; the handle shell, the connection pipe and the support base are sequentially connected and communicate; the first clip arm or the second clip arm is fixedly connected with an upper side or a lower side of the mounting window; the second clip arm or the first clip arm is movably connected with the lower side or the upper side of the mounting window.

Abstract: An auricle clamp for heart surgery comprises: two clamp arms two springs located at two ends of the clamp arms. The two springs are configured to: connect the two ends of the two clamp arms to form a closed-loop structure; and provide a clamping force that enables closing of the two clamp arms; wherein: a clamp arm of the two clamp arms comprises a first rod-shaped tube segment and a second rod-shaped tube segment joined at a hinge site; in a first state, the two clamp arms are substantially parallel; and in a second state, the two clamp arms form a quadrilateral shape, wherein the first rod-shaped tube segments and the second rod-shaped tube segments of the two clamp arms form a majority of the quadrilateral shape.

Abstract: Disclosed are an auricle clamp (101) and a delivery device thereof. The auricle clamp (101) includes two parallel clamp arms (202, 302); and two springs (201, 301) located at two ends of the parallel clamp arms (202, 302) respectively and are used for connecting the two ends of the two clamp arms (202, 302) to form a closed-loop structure and providing a clamping force enabling the two clamp arms (202, 203) to be located close to each other; and at least one of the two clamp arms (202, 302) is divided into a first segment and a second segment, and the first segment and the second segment are pivotally connected or connected via a flexible material.

Abstract: An optical module may include an optical subassembly having a receptacle. The receptacle may have a first diameter. The optical module may include a housing having a circular opening for receiving the receptacle. The circular opening may have a second diameter. The first diameter and the second diameter may be sized to reduce electromagnetic interference at a cut-off frequency from the optical module. The cut-off frequency may be defined by a data rate of at least one component of the optical module.

Abstract: The present disclosure discloses a knot-tying device and a knot-tying system with the knot-tying device. The knot-tying device comprises a shear knife provided at a front end of a driving rod. Hilts of the shear knife (18) are opened by an elastic member, and the hilts are pressed by the driving rod to close and cut a suture. The shear knife is used as a substitute for a cutter in the related art, and is driven to close by the driving rod so as to realize a shearing function. When the suture is cut off by the shear knife, a requirement for sharpness of blades are greatly reduced, and the blades are able to be realized by a common processing method, thereby greatly reducing cost while ensuring reliability. In addition, abrasion on the blades from the suture is reduced, thereby greatly prolonging the service life of the product.

Abstract: A system for repairing a valve leaflet in minimally invasive surgery includes: a delivery device comprising: an operating handle; an outer delivery tube coupled to the operating handle; a delivery needle coupled to the operating handle; and a pushing tube coupled to the operating handle. The system further comprises a repair component comprising an anchor and a connection wire. The delivery device is configured to perform, in response to a trigger operation, a linkage release operation that includes actuating the delivery needle and releasing the repair component.

Abstract: Aspects of the subject disclosure may include, for example, a device comprising: a first micro-camera-element comprising a first sensor area and a first aperture element, the first aperture element having a first structural configuration, the first aperture element and the first sensor area being disposed relative to each other in order to cooperate in obtaining first imaging data having first characteristics, and the first characteristics comprising first imaging resolution and first angular coverage; a second micro-camera-element comprising a second sensor area and a second aperture element, the second aperture element having a second structural configuration, the second aperture element and the second sensor area being disposed relative to each other in order to cooperate in obtaining second imaging data having second characteristics, the second characteristics comprising second imaging resolution and second angular coverage, and the first imaging resolution differing from the second imaging resolution, th