Patents by Inventor Ruizhen HU
Ruizhen HU 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).
-
Publication number: 20230011757Abstract: A method for generating a strategy of an object transport-and-pack process, and a computer device. The method includes: obtaining a priority diagram of a to-be-packed object set in an initial container; encoding the priority diagram; performing a convolution calculation by a neural network, and determining a current to-be-packed object; generating a packing sequence of to-be-packed objects in the to-be-packed object set; and determining a strategy of a packing process of the to-be-packed object set according to a packing target, the packing sequence of the to-be-packed objects, and a transport state of each to-be-packed object and a preset layout strategy.Type: ApplicationFiled: September 19, 2022Publication date: January 12, 2023Inventors: Ruizhen HU, Hui HUANG, Hao ZHANG, Minglun GONG, Juzhan XU, Bin CHEN
-
Publication number: 20220383572Abstract: A floor plan image generating methodology is provided. The methodology includes: acquiring a boundary of a target building and a layout constraint of the target building; outputting multiple first floor plan images according to the layout constraint of the target building; selecting multiple second floor plan images from the multiple first floor plan images; applying a layout constraint of each of the second floor plan images to the boundary of the target building, and obtaining a layout of the target building corresponding to each of the multiple second floor plan images; inputting the layout of the target building and the boundary of the target building into a floor plan image generating network; and obtaining a predicted floor plan image of the target building outputted by the floor plan image generating network.Type: ApplicationFiled: August 10, 2022Publication date: December 1, 2022Inventors: Ruizhen HU, Hui HUANG, Hao ZHANG, Zeyu HUANG, Yuhan TANG
-
Patent number: 11403812Abstract: A three-dimensional (3D) object reconstruction method a computer apparatus and a storage medium. A terminal acquires a scanning image sequence of a target object with the scanning image sequence including at least one frame of at least one scanning image including depth information, uses a neural network algorithm to acquire a predicted semantic label of each scanning image based on the at least one scanning image in the scanning image sequence, and then reconstructs a 3D model of the target object according to the at least one predicted semantic labels and the at least one scanning image in the scanning image sequence. In one aspect, the reconstructing comprises: mapping, according to label distribution corresponding to each voxel of a 3D preset model and each label of the at least one predicted semantic label, each scanning image to a corresponding position of the 3D preset model, to obtain the 3D model.Type: GrantFiled: November 9, 2018Date of Patent: August 2, 2022Assignee: SHENZHEN UNIVERSITYInventors: Ruizhen Hu, Hui Huang
-
Patent number: 11348304Abstract: A posture prediction method includes: acquiring an action area in the scene image by acquiring a scene image obtained by scanning a three-dimensional scene, acquiring an action label corresponding to the action area, the action label is configured to represent the action performed by the human body in the action area, acquiring posture data according to the action area and the action label, acquiring a human body posture model according to the posture data, and mapping the human body posture model into the three-dimensional scene according to the action area and the posture data.Type: GrantFiled: May 9, 2019Date of Patent: May 31, 2022Assignee: SHENZHEN UNIVERSITYInventors: Ruizhen Hu, Hui Huang, Hao Zhang
-
Patent number: 11288538Abstract: A method is disclosed. The method includes obtaining an object for prediction and a plurality of candidate scenes by a computer device; inputting the object for prediction and a current candidate scene to a distance measurement model, the distance measurement model calculates a feature vector corresponding to the current candidate scene based on a trained scene feature subnetwork, and outputs a distance from the object for prediction to the current candidate scene based on the object for prediction and the feature vector corresponding to the current candidate scene, model parameters of the distance measurement model including a parameter determined by a trained object feature subnetwork; obtaining distances from the object for prediction to the plurality of candidate scenes based on the distance measurement model; determining a target scene corresponding to the object for prediction based on the distances from the object for prediction to the plurality of candidate scenes.Type: GrantFiled: July 17, 2018Date of Patent: March 29, 2022Assignee: Shenzhen UniversityInventors: Ruizhen Hu, Hui Huang, Hao Zhang
-
Patent number: 11222244Abstract: A method and a device of acquiring an appearance model, a computer device and a storage medium. The method includes acquiring a set of target point clouds, the set of target point clouds including at least two point clouds, each of the point clouds being obtained by sampling a local area of a target object; then performing a registration and a completion for the at least two point clouds to obtain a processed point cloud, wherein the registration is to transform different point clouds into a same coordinate system, and the completion is to add points in other areas of the object according to existing points in a point cloud to be completed; and finally acquiring the appearance model of the target object according to the processed point cloud. The method can improve applicability of the appearance model of the target object acquired only by registration technique.Type: GrantFiled: July 4, 2020Date of Patent: January 11, 2022Assignee: Shenzhen UniversityInventors: Hui Huang, Zihao Yan, Ruizhen Hu
-
Publication number: 20210365718Abstract: A method is disclosed. The method includes obtaining an object for prediction and a plurality of candidate scenes by a computer device; inputting the object for prediction and a current candidate scene to a distance measurement model, the distance measurement model calculates a feature vector corresponding to the current candidate scene based on a trained scene feature subnetwork, and outputs a distance from the object for prediction to the current candidate scene based on the object for prediction and the feature vector corresponding to the current candidate scene, model parameters of the distance measurement model including a parameter determined by a trained object feature subnetwork; obtaining distances from the object for prediction to the plurality of candidate scenes based on the distance measurement model; determining a target scene corresponding to the object for prediction based on the distances from the object for prediction to the plurality of candidate scenes.Type: ApplicationFiled: July 17, 2018Publication date: November 25, 2021Inventors: Ruizhen HU, Hui HUANG, Hao ZHANG
-
Publication number: 20210327126Abstract: A three-dimensional (3D) object reconstruction method, a device, an apparatus and a storage medium. By acquiring a scanning image sequence of a target object with the scanning image sequence including at least one frame of scanning image including depth information, and adopting a neural network algorithm, a terminal predicts the scanning images in the scanning image sequence, and acquires a predicted semantic label of each scanning image. A 3D model of the target object is then reconstructed according to the predicted semantic labels and the scanning images in the scanning image sequence. The terminal adopts the neural network algorithm to acquire the predicted semantic labels of the scanning images, and a 3D model of the target object is reconstructed according to the predicted semantic labels and the scanning images, so that the reconstructed 3D model is more accurate, and the accuracy of the reconstructed 3D model is improved.Type: ApplicationFiled: November 9, 2018Publication date: October 21, 2021Inventors: Ruizhen HU, Hui HUANG
-
Publication number: 20210279522Abstract: A method and a device of acquiring an appearance model, a computer device and a storage medium. The method includes acquiring a set of target point clouds, the set of target point clouds including at least two point clouds, each of the point clouds being obtained by sampling a local area of a target object; then performing a registration and a completion for the at least two point clouds to obtain a processed point cloud, wherein the registration is to transform different point clouds into a same coordinate system, and the completion is to add points in other areas of the object according to existing points in a point cloud to be completed; and finally acquiring the appearance model of the target object according to the processed point cloud. The method can improve applicability of the appearance model of the target object acquired only by registration technique.Type: ApplicationFiled: July 4, 2020Publication date: September 9, 2021Inventors: Hui HUANG, Zihao Yan, Ruizhen Hu
-
Publication number: 20210248808Abstract: A posture prediction method includes: acquiring an action area in the scene image by acquiring a scene image obtained by scanning a three-dimensional scene, acquiring an action label corresponding to the action area, the action label is configured to represent the action performed by the human body in the action area, acquiring posture data according to the action area and the action label, acquiring a human body posture model according to the posture data, and mapping the human body posture model into the three-dimensional scene according to the action area and the posture data.Type: ApplicationFiled: May 9, 2019Publication date: August 12, 2021Applicant: SHENZHEN UNIVERSITYInventors: Ruizhen HU, Hui HUANG, Hao ZHANG
-
Patent number: 11004206Abstract: The present disclosure provides a three-dimensional shape expression method and device thereof. The method includes following steps: extracting a hybrid type framework of a three-dimensional shape; obtaining a segmentation of the three-dimensional shape by segmenting the hybrid type framework; obtaining a sub-structure of the three-dimensional shape according to the segmentation of the three-dimensional shape; and establishing an expression of the three-dimensional shape by using a bag-of-words model according to the sub-structure of the three-dimensional shape. The embodiments of the present disclosure are capable to express a three-dimensional shape with an easy and high-efficient method.Type: GrantFiled: March 21, 2017Date of Patent: May 11, 2021Assignee: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCESInventors: Wenchao Li, Ruizhen Hu, Hui Huang
-
Patent number: 10664976Abstract: A method for part mobility prediction based on a static snapshot consisting of several steps: constructing mobility units from each 3D model in a set of 3D models with parts segmented according to their motion and grouping all of the mobility units according to their motion type; computing snapshot descriptors for every static snapshot in a mobility unit; learning a snapshot-to-unit distance measure for every motion types of the mobility units; getting the most similar mobility unit and its motion type for a query static snapshot by using the snapshot-to-unit distance measure to select with the minimum distance value; generating multiple motion candidates for a query static snapshot according to the achieved mobility unit and its motion type, sampling the generated motion and getting the best motion parameter for the query static snapshot. This invention can predict the part mobility from a static snapshot of an object.Type: GrantFiled: January 16, 2018Date of Patent: May 26, 2020Assignee: Shenzhen UniversityInventors: Ruizhen Hu, Hui Huang
-
Publication number: 20200027215Abstract: The present disclosure provides a three-dimensional shape expression method and device thereof. The method includes following steps: extracting a hybrid type framework of a three-dimensional shape; obtaining a segmentation of the three-dimensional shape by segmenting the hybrid type framework; obtaining a sub-structure of the three-dimensional shape according to the segmentation of the three-dimensional shape; and establishing an expression of the three-dimensional shape by using a bag-of-words model according to the sub-structure of the three-dimensional shape. The embodiments of the present disclosure are capable to express a three-dimensional shape with an easy and high-efficient method.Type: ApplicationFiled: March 21, 2017Publication date: January 23, 2020Inventors: WENCHAO LI, RUIZHEN HU, HUI HUANG
-
Publication number: 20190066308Abstract: This invention proposes a method for part mobility prediction based on a static snapshot. This part mobility prediction method consists of several steps: constructing mobility units from each 3D model in a set of 3D models with parts segmented according to their motion and grouping all of the mobility units according to their motion type; computing snapshot descriptors for every static snapshot in a mobility unit; learning a snapshot-to-unit distance measure for every motion types of the mobility units; getting the most similar mobility unit and its motion type for a query snapshot by using the snapshot-to-unit distance measure to select with the minimum distance value; generating multiple motion candidates for a query static snapshot according to the achieved mobility unit and its motion type, sampling the generated motion and getting the best motion parameter for the query snapshot. This invention can predict the part mobility from a static snapshot of an object.Type: ApplicationFiled: January 16, 2018Publication date: February 28, 2019Applicant: Shenzhen UniversityInventors: Ruizhen HU, Hui HUANG
-
Publication number: 20180089815Abstract: The present invention provides a functionality analysis method and apparatus of given 3D models. This functionality analysis method comprises: computing interaction context for the central object given in each scene; building the correspondence among those scenes based on the computed interaction context; extracting the functional patches on each central object in each scene based on the built correspondence, and forming a set of proto-patches which is a key component of the functionality model; computing a set of geometric features on the sampled points on each consisting functional patch for each proto-patch; learning a regression model from the geometric features on sample points to their weights for each proto-patch; computing the unary and binary features of each functional patch; and refining the feature combination weights to get the final functionality model.Type: ApplicationFiled: May 25, 2017Publication date: March 29, 2018Inventors: Ruizhen HU, Hui HUANG, Hao ZHANG