Zepp Labs, Inc.

Other

Beijing Shunyuan Kaihua Technology Limited

Tsinghua University

Northwestern University

Xiaowei Dai

Zheng Han

Jiangyu Liu

Fan Yang

Yao Zhang

Michael A Peshkin

Guobin Shen

J Edward Colgate

Zeyu Liu

John W Ware

Xianjun Huang

Yusheng Ye

Zhihai Xiang

Qiuhai Lu

Jason Fass

Mi Chen

Nicholas Marchuk

Seong Jae Hwang

Embodiments disclosed provide a solution to detect a rally in a sports game. One or more stroke actions or non-stroke actions are detected based on motion data detected by a sensor attached to a sports instrument of a user. Using a trained stroke classification model, each detected stroke action is classified into a plurality of classes. Additionally, a determination is made whether each detected non-stroke action is an intentional special user action. The determination whether a non-stroke action is an intentional special user action is made based on a customized set of definitions defining one or more special user actions. One or more rallies are then detected based on the classified stroke actions and intentional special user actions.

Automatic rally detection and scoring

A solution is provided for detecting in real time video highlights in a sports video at a mobile computing device. A highlight detection module of the mobile computing device extracts visual features from each video frame of the sports video using a trained feature model and detects a highlight in the video frame based on the extracted visual features of the video frame using a trained detection model. The feature model and detection model are trained with a convolutional neural network on a large corpus of videos to generate category level and pair-wise frame feature vectors. Based on the detection, the highlight detection module generates a highlight score for each video frame of the sports video and presents the highlight scores to users of the computing device. The feature model and detection model are dynamically updated based on the real time highlight detection data collected by the mobile computing device.

Detect sports video highlights for mobile computing devices

A solution is provided for detecting video highlights of a sports video. A video highlight of a sports video is a portion of the sports video and represents a semantically important event captured in the sports video. An audio stream associated with the sports video is evaluated, e.g., the loudness and length of the loudness of the portions of the audio stream. Video segments of the sports video are selected based on the evaluation of the audio stream. Each selected video segment represents a video highlight candidate of the sports video. A trained audio classification model is used to recognize the voice patterns in the audio stream associated with each selected video segment. Based on the comparison of the recognized video patterns with a set of desired voice patterns, one or more video segments are selected as the video highlights of the sports video.

Detect sports video highlights based on voice recognition

A solution is disclosed for conveniently capturing video highlights in a sports event by tracking head rotation movement of a user and synchronizing rotations of a video capturing device (e.g., the user's mobile phone) on a rotary station with the head rotation movement of the user. The user wears a rotation tracking device, which tracks the head rotation movement of the user in connection with watching an event. The head rotation data from the user's head rotation movement is filtered to generate filtered head rotation data, which excludes noise motions in the head rotation data. Corresponding rotary control data is generated based on the filtered head rotation data. The rotary control data is used to instruct a rotary system to rotate synchronously with head rotations of the user when a video capturing device mounted on the rotary system captures videos of the sports event.

Head rotation tracking device for video highlights identification

A touch interface device includes a touch surface configured to be engaged by an object, first and second actuator assemblies operably connected to the touch surface, and a controller operably connected with the first and second actuator assemblies. The first actuator assembly displaces the touch surface in one or more lateral directions along the touch surface at a first frequency. The second actuator assembly displaces the touch surface in an angled direction that is one of at least obliquely or perpendicularly angled to the touch surface at a second frequency. The controller operates the first and second actuator assemblies so that the touch surface varies in engagement with the object to impart a force on the object that is along the touch surface.

Touch interface device and method for applying lateral forces on a human appendage

A system is provided for real-time single-view action recognition for sports videos based on key pose analysis of the sports videos. A training module of the system trains feature models for a sports action distinctively associated with each sports type using a large corpus of training videos. The trained feature models include a player detector for detecting locations of a player in video frames of a training video, a set of key pose identifiers for identifying distinctive poses of a sports action associated with a type of sports, and a meta classifier for determining a likelihood that the sports action has happened in a sports video based on the key poses analysis. Responsive to an input sports video being received for real-time action recognition, a set of trained feature models associated with the sports type of the input video are selected and applied to the input video.

Real-time single-view action recognition based on key pose analysis for sports videos

A solution is provided for object tracking in a sports video is disclosed. A determination whether a position of the object was identified in a previous video frame is made. If the position of the object was identified in the previous video frame, a new position of the object is identified in a current video frame based on the identified position of the object in the previous video frame. An expected position of the object is identified based on the identified position of the object in the previous video frame and a trained object classification model. A determination is made whether the new position is consistent with the expected position; if the new position is consistent with the expected position, the new position is stored as the position of the object in the current frame.

Fast object tracking framework for sports video recognition

The present invention relates to a structural damage detection method for detecting damage of a structure. The method comprises scanning the surface of the structure with a tapper tapping with a tapping force at a predetermined frequency band while scanning; measuring the response signals transmitted from the surface of the structure; transforming the measured signals to acquire the signal spectrum at each position of the structure surface; intercepting from the signal spectrum a spectrum envelope; calculating a damage indicator at each position of the structure surface; and determining a position where the damage indicator changes abruptly as a position of structural damage. The present invention further relates to a structural damage detection system utilizing the structural damage detection method and a structural damage detection device for use therein.

Structural damage detection system, device and method

A data acquisition method and device for motion recognition, a motion recognition system and a computer readable storage medium are disclosed. The data acquisition device for motion recognition comprises: an initial motion recognition module adapted to perform an initial recognition with respect to motion data collected by a sensor and provide motion data describing a predefined range around a motion trigger point to a data storage module for storage; a data storage module adapted to store motion data provided from the initial motion recognition module; and a communications module adapted to forward the motion data stored in the data storage module to a motion computing device for motion recognition. The present invention makes an initial selection to the motion data to be transmitted to the motion computing device under the same sampling rate. Consequently, the present invention reduces pressures on wireless channel transmission and wireless power consumption, and provides high accuracy in motion recognition while providing motion data at the same sampling rate.

Growing community of inventors

Beijing, China

Xiaowei Dai