Patents by Inventor Yefeng Zheng
Yefeng Zheng has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 12361547Abstract: A data processing method includes: acquiring an initial sample angiography image set; performing data expansion processing on a first sample angiography image based on physical characteristics of blood vessels at a target site to obtain a processed sample angiography image, performing label conversion processing on a first label based on the physical characteristics of the blood vessels at the target site to obtain a second label of the processed sample angiography image, and adding the processed sample angiography image and the second label to a target sample angiography image set; and training an angiography image recognition model using the initial sample angiography image set and the target sample angiography image set to obtain a trained angiography image recognition model. The performance of the trained angiography image recognition model is improved by increasing the number of samples.Type: GrantFiled: December 5, 2022Date of Patent: July 15, 2025Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Dong Wei, Yuexiang Li, Yi Lin, Kai Ma, Yefeng Zheng
-
Patent number: 12327358Abstract: This application discloses a method for reconstructing a dendritic tissue in an image performed by a computer device. The method includes: acquiring original image data corresponding to a target image of a target dendritic tissue and corresponding reconstruction reference data determined based on a local reconstruction result of the target dendritic tissue in the target image; applying a target segmentation model to the original image data and the reconstruction reference data to acquire a target segmentation result for indicating a target category of each pixel in the target image, and the target category of any pixel being used for indicating whether the pixel belongs to the target dendritic tissue or not; and reconstructing the target dendritic tissue in the target image based on the target segmentation result to obtain a complete reconstruction result of the target dendritic tissue in the target image.Type: GrantFiled: October 12, 2022Date of Patent: June 10, 2025Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Donghuan Lu, Kai Ma, Yefeng Zheng
-
Patent number: 12322157Abstract: An image classification method includes: performing image segmentation on an unlabeled sample image to obtain image blocks and performing feature extraction on each image block to obtain an initial image feature set including an initial image feature corresponding to each image block, rearranging and combining initial image features in the initial image feature set to obtain a first image feature set and a second image feature set, first image features in the first image feature set and second image features in the second image feature set corresponding to different rearrangement and combination manners, pre-training an image classification model based on the first image feature set and the second image feature set, the image classification model being configured to classify content in an image, and fine-tuning the pre-trained image classification model based on a labeled sample image.Type: GrantFiled: November 30, 2022Date of Patent: June 3, 2025Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Yuexiang Li, Nanjun He, Kai Ma, Yefeng Zheng
-
Patent number: 12322160Abstract: An image classification model training method and apparatus are provided. Classification results of each image outputted by an image classification model are obtained. When the classification results outputted by the image classification model do not meet a reference condition, a reference classification result is constructed based on the classification results outputted by the image classification model. Because the reference classification result can indicate a probability that images belong to each class, a parameter of the image classification model is updated to obtain a trained image classification model based on a total error value between the classification results of the each image and the reference classification result.Type: GrantFiled: October 12, 2022Date of Patent: June 3, 2025Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Donghuan Lu, Junjie Zhao, Kai Ma, Yefeng Zheng
-
Patent number: 12315168Abstract: An image segmentation method includes obtaining target domain images and source domain images, and segmenting the source domain images and the target domain images by using a generative network in a first generative adversarial network. The method further includes segmenting the source domain images and the target domain images by using a generative network in a second generative adversarial network, and determining a first source domain image and a second source domain image according to source domain segmentation losses, and determining a first target domain image and a second target domain image according to target domain segmentation losses. The method also includes performing cross training on the first generative adversarial network and the second generative adversarial network to obtain a trained first generative adversarial network; and segmenting a to-be-segmented image based on the generative network in the trained first generative adversarial network.Type: GrantFiled: January 28, 2022Date of Patent: May 27, 2025Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Luyan Liu, Kai Ma, Yefeng Zheng
-
Patent number: 12288383Abstract: A method for training an image segmentation model includes calling an encoder to perform feature extraction on a sample image and a scale image to obtain a sample image feature and a scale image feature. The method also includes performing a class activation graph calculation to obtain a sample class activation graph and a scale class activation graph. The method also includes calling a decoder to obtain a sample segmentation result of the sample image, and calling the decoder to obtain a scale segmentation result of the scale image. The method also includes calculating a class activation graph loss and calculating a scale loss. The method also includes training the decoder based on the class activation graph loss and the scale loss.Type: GrantFiled: September 29, 2022Date of Patent: April 29, 2025Assignee: Tencent Technology (Shenzhen) Company LimitedInventors: Donghuan Lu, Kai Ma, Yefeng Zheng
-
Patent number: 12249076Abstract: A method for three-dimensional edge detection includes obtaining, for each of plural two-dimensional slices of a three-dimensional image, a two-dimensional object detection result and a two-dimensional edge detection result, stacking the two-dimensional object detection results into a three-dimensional object detection result, and stacking the two-dimensional edge detection results into a three-dimensional edge detection result. The method also includes performing encoding according to a feature map of the three-dimensional image, the three-dimensional object detection result, and the three-dimensional edge detection result, to obtain an encoding result, and performing decoding according to the encoding result, the three-dimensional object detection result, and the three-dimensional edge detection result, to obtain an optimized three-dimensional edge detection result of the three-dimensional image.Type: GrantFiled: March 24, 2022Date of Patent: March 11, 2025Assignee: Tencent Technology (Shenzhen) Company LimitedInventors: Luyan Liu, Kai Ma, Yefeng Zheng
-
Patent number: 12213828Abstract: This application relates to an image data inspection method and apparatus in the field of artificial intelligence (AI) technologies. The method includes obtaining an image to be inspected, the image to be inspected comprising a sequence of slice images; determining a corresponding group of slice images for each target image in the sequence of slice images; extracting a corresponding slice feature map for each slice image in the group of slice images; aligning the slice feature maps extracted corresponding to the group of slice images; aggregating context information of each slice image in the group of slice images by using an aligned feature map; and performing target region inspection on an aggregated feature map, to obtain an inspection result corresponding to the target image, and combining the inspection result corresponding to each target image, to generate an inspection result corresponding to the image to be inspected.Type: GrantFiled: April 15, 2022Date of Patent: February 4, 2025Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Shilei Cao, Hualuo Liu, Yefeng Zheng
-
Publication number: 20250014150Abstract: In an image processing method, style conversion is performed on a sample image by using a generation network, to obtain a reference image. Style recognition is performed on the reference image by using an adversarial network, to determine a style loss between the reference image and the sample image. Image content recognition is performed on the reference image and the sample image, to determine a content loss between the reference image and the sample image. The generation network is trained based on the style loss and the content loss, to obtain a trained generation network.Type: ApplicationFiled: September 20, 2024Publication date: January 9, 2025Applicant: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Xinpeng XIE, Jiawei Chen, Yuexiang Li, Kai Ma, Yefeng Zheng
-
Publication number: 20250005888Abstract: An image processing method includes: obtaining, through a plurality of radio frequency coils, a plurality of pieces of corresponding undersampled frequency-domain data respectively; and performing, by using a plurality of image processing networks that are cascaded, an information supplement operation respectively on the plurality of pieces of frequency-domain data to obtain a plurality of corresponding target restored images, and determining a target reconstructed image based on the plurality of target restored images, a piece of frequency-domain data being configured for obtaining one target restored image, and an image processing network including an image restoring network, a frequency-domain complement network, and a susceptibility estimation network.Type: ApplicationFiled: August 18, 2024Publication date: January 2, 2025Inventors: Ruifen ZHANG, Luyan LIU, Hong WANG, Yawen HUANG, Yefeng ZHENG
-
Patent number: 12171614Abstract: For three-dimensional segmentation from two-dimensional intracardiac echocardiography imaging, the three-dimension segmentation is output by a machine-learnt multi-task generator. Rather than the brute force approach of training the generator from 2D ICE images to output a 2D segmentation, the generator is trained from 3D information, such as a sparse ICE volume assembled from the 2D ICE images. Where sufficient ground truth data is not available, computed tomography or magnetic resonance data may be used as the ground truth for the sample sparse ICE volumes. The generator is trained to output both the 3D segmentation and a complete volume (i.e., more voxels represented than in the sparse ICE volume). The 3D segmentation may be further used to project to 2D as an input with an ICE image to another network trained to output a 2D segmentation for the ICE image. Display of the 3D segmentation and/or 2D segmentation may guide ablation of tissue in the patient.Type: GrantFiled: September 29, 2023Date of Patent: December 24, 2024Assignee: Siemens Medical Solutions USA, Inc.Inventors: Gareth Funka-Lea, Haofu Liao, Shaohua Kevin Zhou, Yefeng Zheng, Yucheng Tang
-
Publication number: 20240412335Abstract: This application provides a method and an apparatus for training an artifact removal model. The method includes obtaining a reference image and a corresponding artifact image; inputting the artifact image into a plurality of sample removal models to obtain artifact removal results corresponding to the artifact image respectively output by the plurality of sample removal models; determining predicted loss values respectively corresponding to the plurality of sample removal models based on pixel differences between the artifact removal results and the reference image; inputting the predicted loss values respectively corresponding to the plurality of sample removal models into a sample weight model to generate weight parameters respectively corresponding to the plurality of predicted loss values; and training the plurality of sample removal models based on the predicted loss values and the weight parameters to obtain an artifact removal model.Type: ApplicationFiled: August 18, 2024Publication date: December 12, 2024Inventors: Hong WANG, Yefeng ZHENG
-
Publication number: 20240412374Abstract: This application provides a training method and apparatus for an image processing model, an electronic device, and a storage medium.Type: ApplicationFiled: August 18, 2024Publication date: December 12, 2024Inventors: Hong LIU, Dong WEI, Donghuan LU, Liansheng WANG, Yefeng ZHENG
-
Publication number: 20240394846Abstract: A training method includes: obtaining a first sample image and at least two types of second sample images; respectively adding at least two damage feature corresponding to the second sample images to the first sample image, to generate at least two types of single degradation images; fusing the single degradation images, to obtain a multiple degradation image corresponding to the first sample image; performing image reconstruction processing on the multiple degradation image, to generate a predicted reconstruction image corresponding to the multiple degradation image; calculating a loss function value based on the second sample images, the single degradation images, the first sample image, and the predicted reconstruction image; and updating a model parameter of the model based on the loss function value.Type: ApplicationFiled: July 29, 2024Publication date: November 28, 2024Inventors: Yawen HUANG, Yefeng ZHENG, LE ZHANG
-
Publication number: 20240355110Abstract: A method for training an image classification model performed by an electronic device and includes: obtaining a plurality of sample source-domain images, a plurality of sample target-domain images, modal tagging results of the sample source-domain images, and category tagging results of the sample source-domain images; determining first category prediction results of the sample source-domain images by using a neural network model; determining first category prediction results of the sample target-domain images by using the neural network model; for a category tagging result, determining a first loss of the category tagging result based on source-domain image feature pairs corresponding to the category tagging result; and training the neural network model based on first losses of category tagging results, the first category prediction results of the sample source-domain images, and the first category prediction results of the sample target-domain images, to obtain an image classification model.Type: ApplicationFiled: June 24, 2024Publication date: October 24, 2024Inventors: Yawen HUANG, Ziyun CAI, Dandan ZHANG, Yuexiang LI, Hong WANG, Yefeng ZHENG
-
Patent number: 12125170Abstract: An image processing method includes obtaining a sample image and a generative adversarial network (GAN), including a generation network and an adversarial network, and performing style conversion on the sample image, to obtain a reference image. The method further includes performing global style recognition on the reference image, to determine a global style loss between the reference image and the sample image, and performing image content recognition on the reference image and the sample image, to determine a content loss between the reference image and the sample image. The method also includes performing local style recognition on the reference image and the sample image, to determine a local style loss of the reference image and a local style loss of the sample image, training the generation network to obtain a trained generation network, and performing style conversion on a to-be-processed image by using the trained generation network.Type: GrantFiled: March 29, 2022Date of Patent: October 22, 2024Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Xinpeng Xie, Jiawei Chen, Yuexiang Li, Kai Ma, Yefeng Zheng
-
Patent number: 12112556Abstract: An image recognition method includes: obtaining a target three-dimensional (3D) image; inputting the target 3D image to a first recognition model; and obtaining the image type of the target 3D image outputted by the first recognition model. The first recognition model is configured to perform image recognition on the target 3D image to obtain an image type of the target 3D image. A convolutional block of the first recognition model is the same as a convolutional block of a second recognition model, and configured to perform image recognition on the target 3D image. The second recognition model is obtained by training an original recognition model using a target training sample, the target training sample including cubes obtained by rotating and sorting N target cubes obtained from a 3D sample image, N being a natural number greater than 1.Type: GrantFiled: August 13, 2021Date of Patent: October 8, 2024Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Xinrui Zhuang, Yuexiang Li, Yefeng Zheng
-
Publication number: 20240312022Abstract: The present disclosure provides methods, devices, apparatus, and storage medium for determining a target image region of a target object in a target image. The method includes: obtaining a target image comprising a target object; obtaining an original mask and an image segmentation model, the image segmentation model comprising a first unit model and a second unit model; downsampling the original mask based on a pooling layer in the first unit model to obtain a downsampled mask; extracting region convolution feature information of the target image based on a convolution pooling layer in the second unit model and the downsampled mask; updating the original mask according to the region convolution feature information; and in response to the updated original mask satisfying an error convergence condition, determining a target image region of the target object in the target image according to the updated original mask.Type: ApplicationFiled: May 24, 2024Publication date: September 19, 2024Applicant: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Yifan HU, Yefeng Zheng
-
Publication number: 20240296567Abstract: Disclosed are a medical image segmentation method and apparatus, a device, a storage medium, and a program product, which relate to the field of artificial intelligence (AI). The method includes: performing image segmentation on a sample medical image through a source domain segmentation model, to obtain a first segmentation result, the source domain segmentation model being obtained through training based on image data in a source domain, the sample medical image being an unannotated image in a target domain; performing image segmentation on the sample medical image through a target domain segmentation model, to obtain a second segmentation result; correcting the first segmentation result based on the second segmentation result and a segmentation confidence level of the target domain segmentation model, to obtain a corrected segmentation result; and updating training on the target domain segmentation model based on the second segmentation result and the corrected segmentation result.Type: ApplicationFiled: May 15, 2024Publication date: September 5, 2024Applicant: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Zhe XU, Donghuan LU, Yefeng ZHENG
-
Patent number: 12056211Abstract: A method for determining a target image to be labeled includes: obtaining an original image and an autoencoder (AE) set, the original image being an image having not been labeled, the AE set including N AEs; obtaining an encoded image set corresponding to the original image by using the AE set, the encoded image set including N encoded images, the encoded images being corresponding to the AEs; obtaining the encoded image set and a segmentation result set corresponding to the original image by using an image segmentation network, the image segmentation network including M image segmentation sub-networks, and the segmentation result set including [(N+1)*M] segmentation results; determining labeling uncertainty corresponding to the original image according to the segmentation result set; and determining whether the original image is a target image according to the labeling uncertainty.Type: GrantFiled: October 14, 2021Date of Patent: August 6, 2024Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITEDInventors: Yifan Hu, Yuexiang Li, Yefeng Zheng