-
21.发明授权公开(公告)号：US11709914B2
公开(公告)日：2023-07-25
申请号：US17510415
申请日：2021-10-26
Applicant: UBTECH ROBOTICS CORP LTD
Inventor： Hanliu Wang , Jun Cheng , Jianxin Pang
IPC: G06V40/16 , G06F18/214 , G06V10/141 , G06V10/94 , G06V10/24 , G06N3/04
CPC classification number: G06F18/2148 , G06N3/04 , G06V10/141 , G06V10/245 , G06V10/95 , G06V40/166 , G06V40/171
Abstract: A backlight face recognition method, a terminal device using the same, and a computer readable storage medium are provided. The method includes: performing a face detection on each original face image in an original face image sample set to obtain a face frame corresponding to the original face image; capturing the corresponding original face images from the original face image sample set, and obtaining a new face image containing background pixels corresponding to the captured original face images from the original face image sample set; preprocessing all the obtained new face images to obtain a backlight sample set and a normal lighting sample set; and training a convolutional neural network using the backlight sample set and the normal lighting sample set until the convolutional neural network reaches a preset stopping condition. The trained convolutional neural network will improve the accuracy of face recognition in complex background and strong light.
-
Patent Agency Ranking
22.发明申请公开(公告)号：US20220207913A1
公开(公告)日：2022-06-30
申请号：US17562963
申请日：2021-12-27
Applicant: UBTECH ROBOTICS CORP LTD
Inventor： Yusheng Zeng , Jun Cheng , Jianxin Pang
Abstract: A method for training a multi-task recognition model includes: obtaining a number of sample images, wherein some of the sample images are to provide feature-independent facial attributes, some of the sample images are to provide feature-coupled facial attributes, and some of the sample images are to provide facial attributes of face poses; training an initial feature-sharing model based on a first set of sample images to obtain a first feature-sharing model; training the first feature-sharing model based on the first set of sample images and a second set of sample images to obtain a second feature-sharing model with a loss value less than a preset second threshold; obtaining an initial multi-task recognition model by adding a feature decoupling model to the second feature-sharing model; and training the initial multi-task recognition model based on the sample images to obtain a trained multi-task recognition model.
-
公开(公告)号：US11373443B2
公开(公告)日：2022-06-28
申请号：US17105667
申请日：2020-11-27
Applicant: UBTECH ROBOTICS CORP LTD
Inventor： Yue Wang , Jun Cheng , Yepeng Liu , Yusheng Zeng , Jianxin Pang , Youjun Xiong
Abstract: The present disclosure provides a method and an apparatus for face recognition and a computer readable storage medium. The method includes: inputting a to-be-recognized blurry face image into a generator of a trained generative adversarial network to obtain a to-be-recognized clear face image; inputting the to-be-recognized clear face image to the feature extraction network to obtain a facial feature of the to-be-recognized clear face image; matching the facial feature of the to-be-recognized clear face image with each user facial feature in a preset facial feature database to determine the user facial feature best matching the to-be-recognized clear face image as a target user facial feature; and determining a user associated with the target user facial feature as a recognition result. Through this solution, the accuracy of the recognition of blurry faces can be improved.
-
公开(公告)号：US20220156534A1
公开(公告)日：2022-05-19
申请号：US17389380
申请日：2021-07-30
Applicant: UBTECH ROBOTICS CORP LTD
Inventor： Yonghui Cai , Jun Cheng , Jianxin Pang , Youjun Xiong
Abstract: A target object detection model is provided. The target object detection model includes a YOLOv3-Tiny model. Through the target object detection model, low-level information in the YOLOv3-Tiny sub-model can be merged with high-level information therein, so as to fuse the low-level information and the high-level information. Since the low-level information can be further used, the comprehensiveness of target detection is effectively improved, and the detection effect of small targets is improved.
-
公开(公告)号：US20220067354A1
公开(公告)日：2022-03-03
申请号：US17463500
申请日：2021-08-31
Applicant: UBTECH Robotics Corp Ltd
Inventor： Chi Shao , Miaochen Guo , Jun Cheng , Jianxin Pang
Abstract: A dynamic gesture recognition method includes: performing detection on each frame of image of a video stream using a preset static gesture detection model to obtain a static gesture in each frame of image of the video stream; in response to detection of a change of the static gesture from a preset first gesture to a second gesture, suspending the static gesture detection model and activating a preset dynamic gesture detection model; and performing detection on multiple frames of images that are pre-stored in a storage medium using the dynamic gesture detection model to obtain a dynamic gesture recognition result.
-
公开(公告)号：US20210166416A1
公开(公告)日：2021-06-03
申请号：US17107860
申请日：2020-11-30
Applicant: UBTECH ROBOTICS CORP LTD
Inventor： Shuping Hu , Jun Cheng , Jingtao Zhang , Miaochen Guo , Dong Wang , Jianxin Pang , Youjun Xiong
Abstract: A robot climbing control method is disclosed. The method obtains an RGB color image and a depth image of stairs, extracts an outline of a target object of a target step on the stairs from the RGB color image, determines relative position information of the robot and the target step according to the depth image and the outline of the target object, and controls the robot to climb the target step according to the relative position information. The embodiment of the present disclosure allows the robot to effectively adjust postures and forward directions on any size of and non-standardized stairs and avoids the deviation of the walking direction, thereby improving the effectiveness and safety of the stair climbing of the robot.
-
-
-
-
-
Search Results
26
records
(took 0.015 seconds)
