Abstract: Provided are image processing method and apparatus for processing an input image using a convolutional neural network system, and a computer-readable storage medium. The convolutional neural network system that includes an input layer, an intermediate layer and an output layer, the image processing method includes: receiving the input image via the input layer; extracting image features of the input image via the intermediate layer; and outputting processing results for the input image via the output layer, wherein the intermediate layer includes at least one network block each of which includes a first convolutional layer, a first grouping rearrangement layer and a second convolutional layer that are cascaded.

Abstract: The present disclosure provides a method, an electronic device, and a medium for image fusion. The method includes calculating a fusion coefficient image M based on a first frame image I1 or based on both the first frame image I1 and a second frame image I2; calculating a first gradient D1 of the first frame image I1 and a second gradient D2 of the second frame image I2; calculating a preliminary fusion result J based on the image M, the first gradient D1 and the second gradient D2; and obtaining an output image I3 based on the image M, the first gradient D1, the second gradient D2 and the preliminary fusion result J, wherein brightness of the first frame image I1 is greater than brightness of the second frame image I2, and wherein the image M is used to mark fusion weights of pixels in the first frame image I1.

Abstract: The application provides a liveness detection method and a liveness detection system. The liveness detection method includes: obtaining first and second face image data of an object to be detected, and at least one of the first and the second face image data being a depth image; determining a first face region and a second face region, determining whether the first and the second face regions correspond to each other, and extracting, when it is determined that the first and the second face region corresponds to each other, a first and a second face image from the first and the second face region respectively; determining a first classification result for the extracted first face image and a second classification result for the extracted second face image; and determining, based on the first classification result and the second classification result, a detection result for the object to be detected.

Abstract: The disclosure provides method for face-to-unlock, authentication device, and non-volatile storage medium. The method for face-to-unlock includes: obtaining to-be-verified expression data of a to-be-verified expression of a to-be-verified object; comparing the to-be-verified expression data with pre-stored expression data to obtain a comparison result; and in the case where the comparison result indicates that the to-be-verified expression data matches the pre-stored expression data, releasing the lock.

Abstract: The application provides a data update method for face-to-unlock authentication, an authentication device and system, and a non-volatile storage medium. The data update method for face-to-unlock authentication includes: acquiring facial image data of a user to be unlock-authenticated; obtaining a plurality of candidate sets from the facial image data; determining whether there is target data satisfying a predetermined condition among the plurality of candidate sets; in the case where it is determined that there is target data satisfying the predetermined condition among the plurality of candidate sets, updating facial image data in a first set by using the target data, wherein a first unlock authentication is performed based on the facial image data in the first set.

Abstract: A facial image generating method, a facial image generating apparatus, and a facial image generating device are proposed. The method comprises: linking an M-dimensional facial feature vector with an N-dimensional demanded feature vector to generate a synthesized feature vector; and generating a synthesized facial image by use of a deep convolutional network for facial image generation and based on the synthesized feature vector. The method further comprises: generating a demand satisfaction score based on the synthesized facial image and the demanded feature vector by use of a deep convolutional network for demand determination; and updating parameters of the deep convolutional network for facial image generation and the deep convolutional network for demand determination based on the demand satisfaction score. A facial image is generated based on a facial feature vector and a demanded feature vector, a facial image with a specific feature can be generated without using the three-dimensional model.

Abstract: A method for face liveness detection and a device for face liveness detection. The method for face liveness detection includes: performing an illumination liveness detection and obtaining an illumination liveness detection result; and determining whether or not a face to be verified passes the face liveness detection at least according to the illumination liveness detection result. Performing of the illumination liveness detection and obtaining of the illumination liveness detection result includes: acquiring a plurality of illumination images of the face to be verified, in which the plurality of illumination images are captured in a process of dynamically changing mode of illumination light irradiated on the face to be verified and are respectively corresponding to various modes of the illumination light; and obtaining the illumination liveness detection result according to a light reflection characteristic of the face to be verified in the plurality of illumination images.

Abstract: The embodiments of the present invention provide training and construction methods and apparatus of a neural network for object detection, an object detection method and apparatus based on a neural network and a neural network. The training method of the neural network for object detection, comprises: inputting a training image including a training object to the neural network to obtain a predicted bounding box of the training object; acquiring a first loss function according to a ratio of the intersection area to the union area of the predicted bounding box and a true bounding box, the true bounding box being a bounding box of the training object marked in advance in the training image; and adjusting parameters of the neural network by utilizing at least the first loss function to train the neural network.

Abstract: Provided are neural network-based image processing method and apparatus, and a computer-readable storage medium. The image processing method includes: inputting an image into an optimized neural network; extracting, by the optimized neural network, image features of the image; and outputting the image features, wherein the optimized neural network is obtained by performing a first optimization process on at least one sub-layer in a pre-trained initial neural network, each sub-layer of the at least one sub-layer includes a convolutional layer, and the first optimization process comprises: for each sub-layer of the at least one sub-layer, determining one or more channels to be removed from a filter of the convolutional layer and removing said one or more channels, and optimizing parameters of remaining channels in the filter of the convolutional layer, so that error of output features of each optimized sub-layer is minimized.

Abstract: This application provides a video monitoring method and process. The video monitoring method comprises: obtaining first and second video data of a scene being monitored; detecting at least one target object based on the first video data, and determining parameter information of the at least one target object in at least one frame of the first video data, the parameter information including a first position; determining, based on coordinate transforming relationship between the first and the second video data, a second position of the at least one target object in a corresponding frame in the second video data according to the first position; and extracting, based on the second video data, feature information of the at least one target object located in the second position, wherein orientations with which the first and the second video acquiring module acquire video data with respect to a ground plane are different.

Abstract: This application provides a target detection method and device based on a neural network and training method and device of a neural network for target detection. The target detection method comprises: acquiring a to-be-detected image that contains a target; acquiring first feature information of the to-be-detected image by use of a first neural network, acquiring second feature information of the to-be-detected image by use of a second neural network; combining the first feature information and the second feature information to acquire combined feature information; and acquiring a target detection result by use of the second neural network and based on the combined feature information, wherein the number of layers of the second neural network is larger than the number of layers of the first neural network, the first feature information is heatmap feature information, and the second feature information is picture feature information.

Abstract: The present invention relates to a living body detection method and living body detection system capable of realizing living body detection of a human body, and a computer program product. The living body detection method comprises: respectively using each light source of at least two light sources arranged at different positions to illuminate a face of an object to be detected; capturing a plurality of images of the face of the object to be detected that is illuminated by each light source; calculating a difference image between the plurality of images; and obtaining a detection value of the difference image, and determining that the object to be detected is a living body if the detection value is greater than a pre-determined threshold value.

Abstract: The present invention relates to a living body detection method, a living body detection system and a computer program product, which can realize living human body detection. The living body detection method includes: using a laser light source to irradiate the face of an object to be detected; capturing an image of the face of the object to be detected that is irradiated by the laser light source; calculating a light spot area of the image of the face of the object to be detected; and comparing the light spot area with a first predetermined area threshold value, and if the light spot area is greater than the first predetermined area threshold value, determining that the object to be detected is a living body.

Abstract: Provided is a facial feature adding method, a facial feature adding apparatus, and a facial feature adding device. The facial feature adding method comprises: generating an image to be superimposed based on a given facial image and a feature to be added on the given facial image; and superimposing the image to be superimposed and the given facial image to generate a synthesized facial image. In addition, the facial feature adding method further comprises: generating a first face satisfaction score and a second face satisfaction score by use of a deep convolutional network for face determination and based on the synthesized facial image and a real image with the feature to be added; calculating an L1 norm of the image to be superimposed; and updating parameters of networks based on the first face satisfaction score, the second face satisfaction score, and the L1 norm.

Abstract: This application provides a video monitoring method and device. The video monitoring method includes: obtaining video data; inputting at least one frame in the video data into a first neural network to determine object amount information of each pixel dot in the at least one frame; and executing at least one of the following operations by using a second neural network: performing a smoothing operation based on the object amount information in the at least one frame to rectify the object amount information; determining object density information of each pixel dot in the at least one frame based on scene information and the object amount information; predicting object density information of each pixel dot in a to-be-predicted frame next to the at least one frame based on the scene information, the object amount information, and association information between the at least one frame and the to-be-predicted frame.

Abstract: An object detection method and an object detection apparatus are provided. The object detection method includes: mapping at least one image frame in an image sequence into a three dimensional physical space to obtain three dimensional coordinates of each pixel in the at least one image frame; extracting a foreground region in the at least one image frame; segmenting the foreground region into a set of blobs; and detecting, for each blob in the set of blobs, an object in the blob through a neural network based on the three dimensional coordinates of at least one predetermined reference point in the blob, to obtain an object detection result.