Patents Assigned to Shenzhen Horizon Robotics Technology Co., Ltd.
-
Publication number: 20240005917Abstract: Disclosed are a speech interaction method, computer readable storage medium, and electronic device, improving processing efficiency in multi-channel speech interaction scenarios with low resource consumption and processing delay. The method includes: obtaining at least one channel of audio signal; recognizing the at least one channel of audio signal using a preset speech recognition model, to obtain a first type of recognition result; determining stored recognition data from a buffer; generating a second type of recognition result based on the stored recognition data; processing the first and second type of results using the model, to obtain at least one statement recognition result corresponding to the at least one channel of audio signal; performing semantic parsing on the statement recognition result, to obtain at least one parsing result; and generating, based on the at least one parsing result, an instruction for controlling a speech interaction device to perform a corresponding function.Type: ApplicationFiled: February 16, 2022Publication date: January 4, 2024Applicant: Shenzhen Horizon Robotics Technology Co., Ltd.Inventors: Chuan TIAN, Fuping PAN, Jianwei NIU, Kai YU
-
Publication number: 20240004606Abstract: Disclosed in the embodiments of the present disclosure are an audio playback method and apparatus, a computer readable storage medium, and an electronic device. The method includes: acquiring intention determination data collected for at least one user within a target space; determining that at least one user has a target vocal intention based on the intention determination data, and then determining feature information representing a current feature of the at least one user; and extracting and playing an audio corresponding to the feature information from a preset audio library. According to the embodiments of the present disclosure, it is achieved to automatically perform a determination on the target vocal intention of the user by the electronic device, without triggering an audio playing operation by the user, the steps of performing the audio playing operation by the user being omitted, and the convenience of the audio playing operation being improved.Type: ApplicationFiled: February 14, 2022Publication date: January 4, 2024Applicant: Shenzhen Horizon Robotics Technology Co., Ltd.Inventors: Changbao ZHU, Jianwei NIU, Kai YU
-
Patent number: 11580757Abstract: Disclosed are a fatigue state detection method and apparatus, a medium and a device. The method includes: obtaining image blocks containing an organ area of a target object from a plurality of video frames collected by a camera apparatus disposed in a mobile device, to obtain an image-block sequence that is based on the organ area; determining a fatigue state type of the target object based on the image-block sequence of the organ area; sending the image-block sequence to a cloud server if the fatigue state type meets a first preset type, and rendering the cloud server to detect a fatigue level of the target object based on the image-block sequence; and receiving fatigue level information about the target object that is returned by the cloud server. The present disclosure may improve accuracy of fatigue state detection, thereby helping to improve driving safety of the mobile device.Type: GrantFiled: November 17, 2021Date of Patent: February 14, 2023Assignee: Shenzhen Horizon Robotics Technology Co., Ltd.Inventors: Cong Yang, Zhenyu Yang, Weiyu Li
-
Publication number: 20220198331Abstract: Disclosed are a machine model update method and apparatus, a medium, and a device. The method includes: obtaining first hard example samples of a machine model and attribute information of the first hard example samples; determining category-based data distribution information of the first hard example samples according to the attribute information of the first hard example samples; determining second hard example samples of the machine model in current machine learning according to learning requirement information of the machine model for categories of the first hard example samples and the data distribution information; performing, according to learning operation information corresponding to the machine model, a learning operation on the machine model by using the second hard example samples; and updating the machine model based on a learning result of the machine model. The present disclosure helps efficiently implement a machine learning lifecycle, and reduce the cost of machine learning lifecycle.Type: ApplicationFiled: December 6, 2021Publication date: June 23, 2022Applicant: Shenzhen Horizon Robotics Technology Co., Ltd.Inventors: Cong YANG, Yunhui ZHANG, Xhikai ZHANG
-
Publication number: 20220188537Abstract: Disclosed are a behavior recognition method and apparatus, a medium and device. The method includes: obtaining a sequence of image blocks containing a target part of a target object from a plurality of video frames obtained by a camera apparatus disposed in a mobile device; extracting a feature vector on a first preset behavior from the sequence to obtain a first feature vector corresponding to the sequence; determining a behavior type to which a behavior of the target object belongs based on the first feature vector; and if the behavior type to which the behavior belongs satisfies a first behavior type, determining the behavior based on a result of behavior recognition on a second preset behavior for the sequence. The disclosure may realize cascade recognition of the behavior of the target object, thereby helping to improve accuracy of behavior recognition and further driving safety of the mobile device.Type: ApplicationFiled: December 6, 2021Publication date: June 16, 2022Applicant: Shenzhen Horizon Robotics Technology Co., Ltd.Inventors: Yunhui ZHANG, Zhikai Zhang, Cong Yang
-
Publication number: 20220164987Abstract: Disclosed are an extrinsic camera parameter calibration method, apparatus and system. The method includes: obtaining a calibration image photographed for a target location in a calibration cabin by a first camera provided therein; determining calibration feature information of the target location from the calibration image; obtaining reference feature information pre-determined from a reference image photographed for a target location in a reference cabin by a second camera provided therein; determining position-posture change data of the first relative to the second camera based on a calibration location of the calibration feature information in the calibration image and a reference location of the reference feature information in the reference image; and determining an extrinsic parameter of the first camera based on the position-posture change data. The extrinsic parameter of the first camera can be calibrated in real time and is closer to a currently actual state of the first camera when used.Type: ApplicationFiled: November 24, 2021Publication date: May 26, 2022Applicant: Shenzhen Horizon Robotics Technology Co., Ltd.Inventors: Ye TAO, Hongzhi ZHANG