Abstract: A computer vision system including: one or more processors; and memory including instructions that, when executed by the one or more processors, cause the one or more processors to: determine a semantic multi-scale context feature and an instance multi-scale context feature of an input scene; generate a joint attention map based on the semantic multi-scale context feature and the instance multi-scale context feature; refine the semantic multi-scale context feature and instance multi-scale context feature based on the joint attention map; and generate a panoptic segmentation image based on the refined semantic multi-scale context feature and the refined instance multi-scale context feature.

Abstract: A system and a method are disclosed for performing video segmentation, including obtaining a plurality of frames from an input video; extracting a plurality of features from the plurality of frames; obtaining query embeddings corresponding to the plurality of features; refining the query embeddings in a forward time order to generate forward embeddings, and in a backward time order to generate backward embeddings; fusing the forward embeddings and the backward embeddings to obtain fused embeddings; and generating a classification prediction corresponding to the input video based on the fused embeddings.

Abstract: A method and system for training a neural network are provided. The method includes receiving an input image, selecting at least one data augmentation method from a pool of data augmentation methods, generating an augmented image by applying the selected at least one data augmentation method to the input image, and generating a mixed image from the input image and the augmented image.

Abstract: Methods and systems for performing video prediction, including obtaining an input frame from among a plurality of frames included in an input video; extracting a first feature map by providing the input frame to a first plurality of feature extraction layers and a first strided convolutional layer included in an encoder; providing the first feature map and at least one neighboring first feature map corresponding to at least one neighboring frame to a first fusion module included in the encoder; fusing the first feature map with the at least one neighboring first feature map to generate a fused first feature map using the first fusion module; generating a prediction corresponding to the input frame based on the fused first feature map using a decoder; and performing a video prediction task using the prediction.

Abstract: A surge protection circuit includes a first surge protection device (SPD) including a first terminal coupled to a voltage potential and a second terminal. A second SPD includes a first terminal coupled to the voltage potential and a second terminal. A diode array includes a first plurality of diodes, each diode of the first plurality of diodes including an anode coupleable to a respective power line of an electronic circuit and a cathode coupled to the second terminal of the first SPD. The diode array also includes a second plurality of diodes, each diode of the second plurality of diodes including a cathode coupled to a respective power line of the electronic circuit and an anode coupled to the second terminal of the second SPD. The second terminal of the first SPD is further coupled to a common power bus of the electronic circuit.

Abstract: A system and a method are disclosed for tuning parameters of a large language model. The method comprises identifying first weights of a machine learning (ML) model. Second weights are received from a client device. The second weights may be based on updating, by the client device, the first weights. An update matrix may be generated based on the second weights. The update matrix may be decomposed into first decomposition matrices. Singular values that satisfy a criterion may be identified based on the first decomposition matrices. Singular vectors may be identified based on the singular values. Second decomposition matrices may be identified based on the singular vectors. Updates may be received from the client device of third weights associated with the second decomposition matrices. An updated ML model may be generated based on the updates of the third weights. An inference may be generated based on the updated ML model.

Abstract: Disclosed are an electrostatically driven comb structure of an MEMS (Micro Electro Mechanical System), a micro-mirror using the same, and a preparation method therefor.

Abstract: Disclosed are an electrostatically driven comb structure of an MEMS (Micro Electro Mechanical System), a micro-mirror using the same, and a preparation method therefor.

Abstract: The disclosure provides a method, apparatus, electronic device and storage medium for data processing. The method includes: obtaining a storage request for target data; obtaining first dimension information and second dimension information of the target data based on the storage request, the first dimension information representing information of a user to which the target data belongs, and the second dimension information representing information of a business to which the target data belongs; and storing the target data into the first storage region.

Abstract: A system and a method to train a neural network are disclosed. A first image is weakly and strongly augmented. The first image, the weakly and strongly augmented first images are input into a feature extractor to obtain augmented features. Each weakly augmented first image is input to a corresponding first expert head to determine a supervised loss for each weakly augmented first image. Each strongly augmented first image is input to a corresponding second expert head to determine a diversity loss for each strongly augmented first image. The feature extractor is trained to minimize the supervised loss on weakly augmented first images and to minimize a multi-expert consensus loss on strongly augmented first images. Each first expert head is trained to minimize the supervised loss for each weakly augmented first image, and each second expert head is trained to minimize the diversity loss for each strongly augmented first image.

Abstract: Disclosed is a method including receiving, in a semantic segmentation network, input data from a plurality of frames, computing a ground truth label on the plurality of frames, generating a ground truth temporal semantic boundary map from the ground truth label on the plurality of frames, generating a predicted temporal semantic boundary map based on an output of the input data, and determining a loss based on the ground truth temporal semantic boundary map and the predicted temporal semantic boundary map.

Abstract: A system and a method are disclosed for processing and combining feature maps using a hardware friendly multi-kernel convolution block (HFMCB). The method including splitting an input feature map into a plurality of feature maps, each of the plurality of feature maps having a reduced number of channels; processing each of the plurality of feature maps with a different series of kernels; and combining the processed plurality of feature maps.

Abstract: Disclosed are an automatic welding system and method for large structural parts based on hybrid robots and 3D vision. The system comprises a hybrid robot system composed of a mobile robot and an MDOF robot, a 3D vision system, and a welding system used for welding. The rough positioning technique based on a mobile platform and the accurate recognition and positioning technique based on high-accuracy 3D vision are combined, so the working range of the MDOF robot in the XYZ directions is expanded, and flexible welding of large structural parts is realized. The invention adopts 3D vision, thus having better error tolerance and lower requirements for the machining accuracy of workpieces, positioning accuracy of mobile robots and placement accuracy of the workpieces; and the cost is reduced, the flexibility is improved, the working range is expanded, labor is saved, production efficiency is improved, and welding quality is improved.

Abstract: A method of making a plasmonic metal/graphene heterostructure comprises heating an organometallic complex precursor comprising a metal at a first temperature T1 for a first period of time t1 to deposit a layer of the metal on a surface of a heated substrate, the heated substrate in fluid communication with the precursor; and heating, in situ, the precursor at a second temperature T2 for a second period of time t2 to simultaneously form on the layer of the metal, a monolayer of graphene and a plurality of carbon-encapsulated metal nanostructures comprising the metal, thereby providing the plasmonic metal/graphene heterostructure. The heated substrate is characterized by a third temperature T3. The plasmonic metal/graphene heterostructures, devices incorporating the heterostructures, and methods of using the heterostructures are also provided.

Abstract: A method for computing a dominant class of a scene includes: receiving an input image of a scene; generating a segmentation map of the input image, the segmentation map being labeled with a plurality of corresponding classes of a plurality of classes; computing a plurality of area ratios based on the segmentation map, each of the area ratios corresponding to a different class of the plurality of classes of the segmentation map; and outputting a detected dominant class of the scene based on a plurality of ranked labels based on the area ratios.

Abstract: Disclosed are an automatic welding system and method for large structural parts based on hybrid robots and 3D vision. The system comprises a hybrid robot system composed of a mobile robot and an MDOF robot, a 3D vision system, and a welding system used for welding. The rough positioning technique based on a mobile platform and the accurate recognition and positioning technique based on high-accuracy 3D vision are combined, so the working range of the MDOF robot in the XYZ directions is expanded, and flexible welding of large structural parts is realized. The invention adopts 3D vision, thus having better error tolerance and lower requirements for the machining accuracy of workpieces, positioning accuracy of mobile robots and placement accuracy of the workpieces; and the cost is reduced, the flexibility is improved, the working range is expanded, labor is saved, production efficiency is improved, and welding quality is improved.

Abstract: A method for computing a dominant class of a scene includes: receiving an input image of a scene; generating a segmentation map of the input image, the segmentation map being labeled with a plurality of corresponding classes of a plurality of classes; computing a plurality of area ratios based on the segmentation map, each of the area ratios corresponding to a different class of the plurality of classes of the segmentation map; and outputting a detected dominant class of the scene based on a plurality of ranked labels based on the area ratios.

Abstract: A method and system for training a neural network are provided. The method includes receiving an input image, selecting at least one data augmentation method from a pool of data augmentation methods, generating an augmented image by applying the selected at least one data augmentation method to the input image, and generating a mixed image from the input image and the augmented image.

Abstract: A computer vision system including: one or more processors; and memory including instructions that, when executed by the one or more processors, cause the one or more processors to: determine a semantic multi-scale context feature and an instance multi-scale context feature of an input scene; generate a joint attention map based on the semantic multi-scale context feature and the instance multi-scale context feature; refine the semantic multi-scale context feature and instance multi-scale context feature based on the joint attention map; and generate a panoptic segmentation image based on the refined semantic multi-scale context feature and the refined instance multi-scale context feature.

Abstract: A method and system for training a neural network are provided. The method includes receiving an input image, selecting at least one data augmentation method from a pool of data augmentation methods, generating an augmented image by applying the selected at least one data augmentation method to the input image, and generating a mixed image from the input image and the augmented image.