Abstract: Fast visual data annotation includes automatic detection using an automatic detector to detect subjects and joints in video frames. Then, annotation with sampling is performed, including determining when a frame is a sample (e.g., based on comparison of frames). Replay and refinement is utilized where user is involved with manually annotating subjects and/or joints in only select video frames.

Abstract: Fast visual data annotation includes automatic detection using an automatic detector to detect subjects and joints in video frames. Then, annotation with sampling is performed, including determining when a frame is a sample (e.g., based on comparison of frames). Replay and refinement is utilized where user is involved with manually annotating subjects and/or joints in only select video frames.

Abstract: A system for marker-based pose estimation includes display of a 3D human body model on a display screen. The system includes circuitry that assigns a weight to each marker of a plurality of markers distributed on a human body. The circuitry determines an objective function for each joint of the plurality of joints based on a set of descendant markers. The circuitry updates each parameter of a first plurality of parameters that defines a first pose of the human body, based on a change in a value of the objective function. The circuitry determines a second plurality of parameters that defines a second pose of the human body based on the updated first plurality of parameters. The circuitry controls movement of the rigged 3D human body model on the display screen in synchronization with the second pose.

Abstract: Various aspects of a system and a method for camera calibration by use of a rotatable three-dimensional (3-D) calibration object are disclosed herein. In accordance with an embodiment, the system includes a first electronic device, which determines a rotation pattern of the 3-D calibration object, based on a set of pre-selected images. The set of pre-selected images includes the 3-D calibration object captured at pre-defined viewing angles. Control information is communicated by the first electronic device to a second electronic device associated with the 3-D calibration object to rotate the 3-D calibration object in accordance with the determined rotation pattern. A plurality of image frames of the 3-D calibration object are captured to calibrate intrinsic and/or extrinsic camera parameters of the first electronic device.

Abstract: Various aspects of a system and a method for extrinsic camera parameters calibration by use of a three dimensional (3D) calibration are disclosed herein. In accordance with an embodiment, the system includes an electronic device, which receives an image, which includes the 3D calibration object with a plurality of unique identifiers. One or more appearance parameters of the 3D calibration object and/or a plurality of intrinsic camera parameters of an image-capturing device that captured the received image, are further received. A positional probability is computed for each of the plurality of unique identifiers distributed in the received image. The plurality of unique identifiers of the 3D calibration object are located in the received image to calibrate the extrinsic camera parameters based on the computed positional probability, the received one or more appearance parameters of the 3D calibration object, and the plurality of intrinsic camera parameters.

Abstract: A method to position multiple camera using camera ordering is described herein. Based on two known camera positions, a third camera position is able to be determined using acquired information and calculations.

Abstract: A method to estimate a set of camera locations, in clockwise or counter-clockwise order, according to the videos captured by these cameras is described herein. In some embodiments, the cameras are assumed to be fixed, with no or very mild tilting angles and no rolling angles (the horizon is horizontal in each camera image). The difference of orientation (rolling angle) between each neighboring (closest) camera pair is able to be up to 45 degrees. Each camera is assumed to have overlapped views with at least one other camera. Each camera has one right neighboring camera and one left neighboring camera, except the first and the last cameras which have only one neighboring camera at one side. The locations of the cameras then are able to be expressed as a unique list counter-clockwise. The input videos are assumed to be synchronized in time.

Abstract: Various aspects of a system and a method for extrinsic camera parameters calibration by use of a three dimensional (3D) calibration are disclosed herein. In accordance with an embodiment, the system includes an electronic device, which receives an image, which includes the 3D calibration object with a plurality of unique identifiers. One or more appearance parameters of the 3D calibration object and/or a plurality of intrinsic camera parameters of an image-capturing device that captured the received image, are further received. A positional probability is computed for each of the plurality of unique identifiers distributed in the received image. The plurality of unique identifiers of the 3D calibration object are located in the received image to calibrate the extrinsic camera parameters based on the computed positional probability, the received one or more appearance parameters of the 3D calibration object, and the plurality of intrinsic camera parameters.

Abstract: Various aspects of a system and a method for camera calibration by use of a rotatable three-dimensional (3-D) calibration object are disclosed herein. In accordance with an embodiment, the system includes a first electronic device, which determines a rotation pattern of the 3-D calibration object, based on a set of pre-selected images. The set of pre-selected images includes the 3-D calibration object captured at pre-defined viewing angles. Control information is communicated by the first electronic device to a second electronic device associated with the 3-D calibration object to rotate the 3-D calibration object in accordance with the determined rotation pattern. A plurality of image frames of the 3-D calibration object are captured to calibrate intrinsic and/or extrinsic camera parameters of the first electronic device.

Abstract: A method to position multiple camera using camera ordering is described herein. Based on two known camera positions, a third camera position is able to be determined using acquired information and calculations.

Abstract: A method to estimate a set of camera locations, in clockwise or counter-clockwise order, according to the videos captured by these cameras is described herein. In some embodiments, the cameras are assumed to be fixed, with no or very mild tilting angles and no rolling angles (the horizon is horizontal in each camera image). The difference of orientation (rolling angle) between each neighboring (closest) camera pair is able to be up to 45 degrees. Each camera is assumed to have overlapped views with at least one other camera. Each camera has one right neighboring camera and one left neighboring camera, except the first and the last cameras which have only one neighboring camera at one side. The locations of the cameras then are able to be expressed as a unique list counter-clockwise. The input videos are assumed to be synchronized in time.

Abstract: The present invention relates to an LED structure having a progressive work function layer, which adopts a conversion layer with a gradually varying work function as the medium for forming an Ohmic contact between the p-type GaN and the metal reflection layer. The work function of the conversion layer is not a single value. Instead, different quantities of dopants are doped at different depths of the conversion layer. Thereby, the conversion layer can match excellently the connected p-type GaN and the metal reflection layer. By taking advantage of the high light transmissivity of the material of the conversion layer, the possibility that light is absorbed by the Ohmic contact layer is reduced. The conversion according to the present invention can also block diffusion of the metal in the metal reflection to the p-type GaN. Accordingly, it can be used as both an Ohmic contact layer and a barrier layer.

Abstract: An p-type organic/p-type inorganic electroluminescent device includes an anode, a p-type organic compound layer connected to the anode, a cathode, and a p-type inorganic compound layer connected to the cathode. The p-type organic compound layer and the p-type inorganic compound layer are in direct contact with one another to form an p-type organic/p-type inorganic hetero-junction therebetween, and thereby light generated at the p-type organic/p-type inorganic hetero-junction with electrons and holes recombination at the hetero-junction under applied forward voltage.

Abstract: An illumination apparatus includes a sunlight guide member, a light emitting device, a wavelength converting element, a fiber-optic light guide connecting the light emitting device and the wavelength converting element, and a lighting control system for controlling the light emitting diode. The sunlight guide member is provided to guide outdoor sunlight to an indoor area. The light emitting device is provided for generating light having a first wavelength range along a beam path. The wavelength converting element is disposed in the beam path and physically separated from the light emitting device. The wavelength converting element converts the light emitted by the light emitting device with the first wavelength range into light having a second wavelength range. The lighting control system is configured to detect an intensity of the outdoor sunlight and to control radiance of the light emitting device according to the detected intensity of the outdoor sunlight.

Abstract: The present invention relates to a method for ternary wafer bonding and the structure thereof. According to the present invention, silver island structures in the second bonding layer are distributed on the first bonding layer deposited on the surface of a single silicon wafer for forming a gold-silver combination structure, which is then bonded with another silicon wafer without any metal layers thereon at a low-temperature thermal process of 250° C. for completing gold-silver-silicon ternary wafer bonding. Thus, the temperature required for gold-silicon bonding is lowered and the process of wafer bonding is simplified as well. In addition, the quality of wafer bonding is also assured.

Abstract: The present invention relates to an LED structure having a progressive work function layer, which adopts a conversion layer with a gradually varying work function as the medium for forming an Ohmic contact between the p-type GaN and the metal reflection layer. The work function of the conversion layer is not a single value. Instead, different quantities of dopants are doped at different depths of the conversion layer. Thereby, the conversion layer can match excellently the connected p-type GaN and the metal reflection layer. By taking advantage of the high light transmissivity of the material of the conversion layer, the possibility that light is absorbed by the Ohmic contact layer is reduced. The conversion according to the present invention can also block diffusion of the metal in the metal reflection to the p-type GaN. Accordingly, it can be used as both an Ohmic contact layer and a barrier layer.

Abstract: Systems and methods are disclosed for identifying the vanishing point, vanishing direction and road width of an image using scene identification algorithms and a new edge-scoring technique.

Abstract: The present disclosure provides a nano-meshed patterned substrate and a method of forming the same. In an embodiment, a metal layer is formed on a substrate, and a heat treatment is performed on the substrate and the metal layer so that the metal layer is transformed into a nano-meshed metal structure. The substrate is then etched using the nano-meshed metal structure as an etch mask. After removing the nano-meshed metal structure, a nano-meshed patterned substrate is obtained.

Abstract: A Sn—Ag bonding and a method thereof are revealed. By means of a bonding layer formed by tin and silver between wafers, the stress released by diffusion and bonding between tin(Sn) and silver(Ag) is larger than the stress released by diffusion and bonding of conventional gold-silver bonding. Moreover, a Sn—Ag bonding method of the present invention forms Sn—Ag bonding at low temperature and releases more stress so as to reduce thermal stress generated during wafer bonding effectively. And after wafer bonding, the high temperature processes can be performed.

Abstract: A Sn—Ag bonding and a method thereof are revealed. By means of a bonding layer formed by tin and silver between wafers, the stress released by diffusion and bonding between tin(Sn) and silver(Ag) is larger than the stress released by diffusion and bonding of conventional gold-silver bonding. Moreover, a Sn—Ag bonding method of the present invention forms Sn—Ag bonding at low temperature and releases more stress so as to reduce thermal stress generated during wafer bonding effectively. And after wafer bonding, the high temperature processes can be performed.