Abstract: Methods and systems for budgeted and simplified training of deep neural networks (DNNs) are disclosed. In one example, a trainer is to train a DNN using a plurality of training sub-images derived from a down-sampled training image. A tester is to test the trained DNN using a plurality of testing sub-images derived from a down-sampled testing image. In another example, in a recurrent deep Q-network (RDQN) having a local attention mechanism located between a convolutional neural network (CNN) and a long-short time memory (LSTM), a plurality of feature maps are generated by the CNN from an input image. Hard-attention is applied by the local attention mechanism to the generated plurality of feature maps by selecting a subset of the generated feature maps. Soft attention is applied by the local attention mechanism to the selected subset of generated feature maps by providing weights to the selected subset of generated feature maps in obtaining weighted feature maps.

Abstract: Provided is a stable formulation of a human papillomavirus virus-like particle vaccine. The stable formulation is composed of a human papillomavirus virus-like particle, a buffer solution, an osmotic pressure regulator, a surfactant and an aluminum adjuvant, wherein the components of the vaccine comprise HPV virus-like particles assembled by L1 proteins of HPV types 6, 11, 16, 18, 31, 33, 45, 52 and 58, and one or more HPV virus-like particles assembled by L1 proteins of other pathogenic HPV types. The formulation can enhance the stability of the vaccine and prolong the validity period of the vaccine in an aqueous formulation.

Abstract: An example apparatus for generating synthesized images includes a receiver to receive a frame, a mask and external images. The apparatus also includes a foreground augmenter to generate augmented foregrounds in the frame based on the mask. The apparatus includes a background augmenter to generate augmented backgrounds based on the frame, the mask, and the external images. The apparatus also further includes an image synthesizer to generate a synthesized image based on the generated augmented foregrounds and the augmented backgrounds.

Abstract: A micro puree machine has a mixing shaft attachable to a blade that is subject to both linear and rotational movement. A drive motor is contained within a lower portion of the machine housing and operatively couples to a first gear that engages a second gear on the mixing shaft for rotation of the mixing blade. A position motor within the upper housing operates to move the mixing blade linearly to engage the ingredients in the bowl. During both axial and rotational movement of the mixing shaft, the drive motor and the position motor remain stationary within their respective positions in the housing.

Abstract: Embodiments of bonded semiconductor structures and fabrication methods thereof are disclosed. In an example, a method for forming a semiconductor device is disclosed. A first device layer is formed on a first substrate. A first bonding layer including a first bonding contact and a first bonding alignment mark is formed above the first device layer. A second device layer is formed on a second substrate. A second bonding layer including a second bonding contact and a second bonding alignment mark is formed above the second device layer. The first bonding alignment mark is aligned with the second bonding alignment mark, such that the first bonding contact is aligned with the second bonding contact. The first substrate and the second substrate are bonded in a face-to-face manner, so that the first bonding contact is in contact with the second bonding contact at a bonding interface, and the first bonding alignment mark is in contact with the second bonding alignment mark at the bonding interface.

Abstract: An object management system (OMS) receives a plurality of objects to be stored. The OMS identifies a subset of objects associated with a same time period and a same namespace from the plurality of objects and generates a batch object comprising the subset of objects. The OMS issues a write request to store the batch object in an object storage system (OSS) and stores, for each object in the subset of objects, a data structure comprising an identifier of the batch object and a position of the object within the batch object. Upon receiving a read request for an object, the OMS determines, based on the data structure, whether the object is stored in the OSS and issues a read request to the OSS based on the identifier of the batch object and the position of the object within the batch, if the objects is stored in the OSS.

Abstract: A Video Semantic Segmentation System (VSSS) is disclosed that performs accurate and fast semantic segmentation of videos using a set of temporally distributed neural networks. The VSSS receives as input a video signal comprising a contiguous sequence of temporally-related video frames. The VSSS extracts features from the video frames in the contiguous sequence and based upon the extracted features, selects, from a set of labels, a label to be associated with each pixel of each video frame in the video signal. In certain embodiments, a set of multiple neural networks are used to extract the features to be used for video segmentation and the extraction of features is distributed among the multiple neural networks in the set. A strong feature representation representing the entirety of the features is produced for each video frame in the sequence of video frames by aggregating the output features extracted by the multiple neural networks.

Abstract: A digital imaging method of analyzing pixel data of an image of a diaper change region associated with an individual for determining an individual-specific diaper change region skin characteristic is provided, including aggregating training images of diaper change regions of individuals; training, with pixel data of the training images, a diaper change region skin model operable to determine diaper change region skin characteristics associated with a diaper change region based on the pixel data of the diaper change region; receiving an image of an individual captured by a digital camera; analyzing, by the diaper change region skin model, the image captured by the digital camera to determine an individual-specific diaper change region skin characteristic; generating an individual-specific recommendation designed to address a feature identifiable within pixel data of the individual's skin based on the individual-specific diaper change region skin characteristic; and providing the individual-specific recommendat

Abstract: Methods and systems for budgeted and simplified training of deep neural networks (DNNs) are disclosed. In one example, a trainer is to train a DNN using a plurality of training sub-images derived from a down-sampled training image. A tester is to test the trained DNN using a plurality of testing sub-images derived from a down-sampled testing image. In another example, in a recurrent deep Q-network (RDQN) having a local attention mechanism located between a convolutional neural network (CNN) and a long-short time memory (LSTM), a plurality of feature maps are generated by the CNN from an input image. Hard-attention is applied by the local attention mechanism to the generated plurality of feature maps by selecting a subset of the generated feature maps. Soft attention is applied by the local attention mechanism to the selected subset of generated feature maps by providing weights to the selected subset of generated feature maps in obtaining weighted feature maps.

Abstract: Methods and systems are disclosed for boosting deep neural networks for deep learning. In one example, in a deep neural network including a first shallow network and a second shallow network, a first training sample is processed by the first shallow network using equal weights. A loss for the first shallow network is determined based on the processed training sample using equal weights. Weights for the second shallow network are adjusted based on the determined loss for the first shallow network. A second training sample is processed by the second shallow network using the adjusted weights. In another example, in a deep neural network including a first weak network and a second weak network, a first subset of training samples is processed by the first weak network using initialized weights. A classification error for the first weak network on the first subset of training samples is determined.

Abstract: Embodiments of methods and structures for forming a 3D integrated wiring structure are disclosed. The method can include forming an insulating layer on a front side of a first substrate; forming a semiconductor layer on a front side of the insulating layer; patterning the semiconductor layer to expose at least a portion of a surface of the insulating layer; forming a plurality of semiconductor structures over the front side of the first substrate, wherein the semiconductor structures include a plurality of conductive contacts and a first conductive layer; joining a second substrate with the semiconductor structures; performing a thinning process on a backside of the first substrate to expose the insulating layer and one end of the plurality of conductive contacts; and forming a conductive wiring layer on the exposed insulating layer.

Abstract: Digital synthesis techniques are described to synthesize a digital image at a target time between a first digital image and a second digital image. To begin, an optical flow generation module is employed to generate optical flows. The digital images and optical flows are then received as an input by a motion refinement system. The motion refinement system is configured to generate data describing many-to-many relationships mapped for pixels in the plurality of digital images and reliability scores of the many-to-many relationships. The reliability scores are then used to resolve overlaps of pixels that are mapped to a same location by a synthesis module to generate a synthesized digital image.

Abstract: In an example, a semiconductor device includes a first semiconductor structure including a memory array device, a second semiconductor structure including a peripheral device, and a bonding structure comprising a first bonding pad, a second bonding pad, and a remainder layer located between and in contact with the first and second bonding pad in a vertical direction. The first bonding pad is located between the remainder layer and the first semiconductor structure in the vertical direction. The second bonding pad is located between the remainder layer and the second semiconductor structure in the vertical direction. A conductive material of the remainder layer is cobalt metal different from the first and second bonding pads.

Abstract: A method and system of multiple facial attributes recognition using highly efficient neural networks.

Abstract: Provided is a stable formulation for a recombinant monoclonal antibody, consisting of a recombinant anti-PD-1 monoclonal antibody, a buffer, an osmotic pressure regulator, a stabilizer, and a surfactant. The pharmaceutical formulation may enhance the stability of the antibody and prolongs a validity period of the antibody in aqueous formulations.

Abstract: Embodiments of bonded semiconductor structures and fabrication methods thereof are disclosed. In an example, a method for forming a semiconductor device is disclosed. A first device layer is formed on a first substrate. A first bonding layer including a first bonding contact is formed above the first device layer. A first capping layer is formed at an upper end of the first bonding contact. The first capping layer has a conductive material different from a remainder of the first bonding contact. A second device layer is formed on a second substrate. A second bonding layer including a second bonding contact is formed above the second device layer. The first substrate and the second substrate are bonded in a face-to-face manner, so that the first bonding contact is in contact with the second bonding contact by the first capping layer.

Abstract: A system for generating scenario objects is described. The system generates, according to one or more scenario expressions executed by the one or more processors, a scenario object comprising an adjustment to a current market data object. In generating the scenario object, the system requests historical market data from a data repository, receives one or more time series conditions, manipulates one or more characteristics of the historical market data based on the one or more time series conditions, and generates, based on the manipulated historical market data, the scenario object. The system applies the scenario object to the current market data object to calculate an adjusted market data object.

Abstract: Embodiments of methods for forming a hybrid-bonded semiconductor structure are disclosed. The method include disposing first second, third, and fourth dielectric layers, forming first and second openings by etching the fourth dielectric layer using a first etching selectivity, etching the third and fourth dielectric layers in the first and second openings respectively using a second etching selectivity, etching the second and third dielectric layers in the first and second openings using the first etching selectivity, etching the first dielectric layer in the first opening and the second dielectric layer in the second opening using the second etching selectivity, etching the first dielectric layer in the first and second openings using the first etching selectivity, and forming conductive material in the first and second openings.

Abstract: The present invention is directed to an aptamer composition comprising at least one oligonucleotide consisting of: deoxyribonucleotides, ribonucleotides, derivatives of deoxyribonucleotides, derivatives of ribonucleotides, and mixtures thereof; wherein said aptamer composition has a binding affinity for one or more bacterial species from the genera Prevotella and Porphyromonas.