Abstract: A bi-directional and multi-channel optical module incudes a casing, an optical transmitter assembly, an optical receiver assembly and an optical fiber adaptor. The optical transmitter assembly is disposed in an accommodation space of the casing. The optical transmitter assembly includes a plurality of light emission units and a wavelength division multiplexer disposed corresponding to the plurality of light emission units. The optical receiver assembly is disposed in the accommodation space. The optical receiver assembly includes a plurality of light receiving units and a wavelength demultiplexer disposed corresponding to the plurality of light receiving units. The optical fiber adaptor is disposed on the casing.

Abstract: A method for manufacturing a semiconductor package structure and a clamp apparatus are provided. The method includes: (a) providing a package body disposed on a chuck, wherein the package body includes at least one semiconductor element encapsulated in an encapsulant; (b) moving a pressing tool transversely to above the package body; and (c) pressing the package body on the chuck through the pressing tool.

Abstract: A method for manufacturing a semiconductor package structure and a semiconductor manufacturing apparatus are provided. The method includes: (a) providing a package body disposed on a chuck, wherein the package body includes at least one semiconductor element encapsulated in an encapsulant; and (b) sucking the package body through the chuck to create a plurality of negative pressures on a bottom surface of the package body sequentially from an inner portion to an outer portion of the package body.

Abstract: A face recognition method includes: extracting a first identity feature of a first face image by using a feature extraction module, and extracting a second identity feature of a second face image by using the feature extraction module, wherein the feature extraction module is implemented by using a neural network, and pre-trained in a manner such that a correlation coefficient of training batch data is obtained based on an identity feature and an age feature of a sample face image in the training batch data, and decorrelated training of the identity feature and the age feature is performed on the feature extraction module based on the correlation coefficient; and performing a face recognition based on determining a similarity between faces in the first face image and the second face image according to the first identity feature and the second identity feature.

Abstract: The disclosure provides an image recognition network model training method, including: acquiring a first image feature corresponding to an image set; acquiring a first identity prediction result by using an identity classifier, and acquiring a first pose prediction result by using a pose classifier; obtaining an identity classifier according to the first identity prediction result and an identity tag, and obtaining a pose classifier according to the first pose prediction result and a pose tag; performing pose transformation on the first image feature by using a generator, to obtain a second image feature corresponding to the image set; acquiring a second identity prediction result by using the identity classifier, and acquiring a second pose prediction result by using the pose classifier; and training the generator.

Abstract: A face recognition method includes: extracting a first identity feature of a first face image by using a feature extraction module, and extracting a second identity feature of a second face image by using the feature extraction module, wherein the feature extraction module is implemented by using a neural network, and pre-trained in a manner such that a correlation coefficient of training batch data is obtained based on an identity feature and an age feature of a sample face image in the training batch data, and decorrelated training of the identity feature and the age feature is performed on the feature extraction module based on the correlation coefficient; and performing a face recognition based on determining a similarity between faces in the first face image and the second face image according to the first identity feature and the second identity feature.

Abstract: A face recognition method includes: extracting a first identity feature of a first face image by using a feature extraction module, and extracting a second identity feature of a second face image by using the feature extraction module, wherein the feature extraction module is implemented by using a neural network, and pre-trained in a manner such that a correlation coefficient of training batch data is obtained based on an identity feature and an age feature of a sample face image in the training batch data, and decorrelated training of the identity feature and the age feature is performed on the feature extraction module based on the correlation coefficient; and performing a face recognition based on determining a similarity between faces in the first face image and the second face image according to the first identity feature and the second identity feature.

Abstract: A method for manufacturing a semiconductor package structure and a clamp apparatus are provided. The method includes: (a) providing a package body disposed on a chuck, wherein the package body includes at least one semiconductor element encapsulated in an encapsulant; (b) moving a pressing tool transversely to above the package body; and (c) pressing the package body on the chuck through the pressing tool.

Abstract: A method for manufacturing a semiconductor package structure and a semiconductor manufacturing apparatus are provided. The method includes: (a) providing a package body disposed on a chuck, wherein the package body includes at least one semiconductor element encapsulated in an encapsulant; and (b) sucking the package body through the chuck to create a plurality of negative pressures on a bottom surface of the package body sequentially from an inner portion to an outer portion of the package body.

Abstract: The disclosure provides an image recognition network model training method, including: acquiring a first image feature corresponding to an image set; acquiring a first identity prediction result by using an identity classifier, and acquiring a first pose prediction result by using a pose classifier; obtaining an identity classifier according to the first identity prediction result and an identity tag, and obtaining a pose classifier according to the first pose prediction result and a pose tag; performing pose transformation on the first image feature by using a generator, to obtain a second image feature corresponding to the image set; acquiring a second identity prediction result by using the identity classifier, and acquiring a second pose prediction result by using the pose classifier; and training the generator.

Abstract: A face recognition method includes: extracting a first identity feature of a first face image by using a feature extraction module, and extracting a second identity feature of a second face image by using the feature extraction module, wherein the feature extraction module is implemented by using a neural network, and pre-trained in a manner such that a correlation coefficient of training batch data is obtained based on an identity feature and an age feature of a sample face image in the training batch data, and decorrelated training of the identity feature and the age feature is performed on the feature extraction module based on the correlation coefficient; and performing a face recognition based on determining a similarity between faces in the first face image and the second face image according to the first identity feature and the second identity feature.

Abstract: The invention provides novel compounds having the general formula: wherein R1, R2, R3, R4, R5, R7, A1, A2 and A3 are as described herein, compositions including the compounds and methods of using the compounds.

Abstract: The invention provides novel compounds having the general formula: wherein R1, R2, R3, R4, R5, R7, A1, A2 and A3 are as described herein, compositions including the compounds and methods of using the compounds.

Abstract: The invention provides novel compounds having the general formula: wherein R1, R2, R3, R4, R5, R7, A1, A2 and A3 are as described herein, compositions including the compounds and methods of using the compounds.

Abstract: An integrated circuit comprises a substrate and a buried dielectric formed in the substrate. The buried dielectric has a first thickness in a first region, a second buried dielectric thickness in a second region, and a step between the first and second regions. A semiconductor layer overlies the buried dielectric.

Abstract: An integrated circuit comprises a substrate and a buried dielectric formed in the substrate. The buried dielectric has a first thickness in a first region, a second buried dielectric thickness in a second region, and a step between the first and second regions. A semiconductor layer overlies the buried dielectric.

Abstract: A method for monitoring and reporting sound pressure level exposure for a user of a first communication device (104) is implemented in one embodiment when the device measures a sound pressure level (SPL) of the surrounding environment. The device stores at least the SPL measurement in a memory, producing an SPL exposure record, and displays a visual representation of the SPL exposure record on a display screen (212). In another embodiment, the SPL is measured by a second communication device (102) and combined with a known SPL for an output audio transducer (306) of the second device, producing a user sound exposure level. The user sound exposure level is transmitted to the first communication device. The user is notified when the user sound exposure level exceeds a predetermined threshold. A server (112) may also be used to track SPLs over time and recommend corrective action when exposure limits are exceeded.

Abstract: A fin-FET device and a method for fabrication thereof both employ a bulk semiconductor substrate. A fin and an adjoining trough are formed within the bulk semiconductor substrate. The trough is partially backfilled with a deposited dielectric layer to form an exposed fin region and an unexposed fin region. A gate dielectric layer is formed upon the exposed fin region and a gate electrode is formed upon the gate dielectric layer. By employing a bulk semiconductor substrate the fin-FET device is fabricated cost effectively.

Abstract: A semiconductor device and method for forming the same including improved electrostatic discharge protection for advanced semiconductor devices, the semiconductor device including providing semiconductor substrate having a pre-selected surface orientation and crystal direction; an insulator layer overlying the semiconductor substrate; a first semiconductor active region overlying the insulator layer having a first surface orientation selected from the group consisting of <100> and <110>; a second semiconductor active region extending through a thickness portion of the insulator layer having a second surface orientation selected from the group consisting of <110> and <100> different from the first surface orientation; wherein MOS devices including a first MOS device of a first conduction type is disposed on the first semiconductor active region and a second MOS device of a second conduction type is disposed on the second semiconductor active region.

Abstract: A photoplethysmographic sensing system for determining a user's pulse rate includes a light emitting device (100, 201, 310, 500) including a first plurality of light emitting particles (108, 208, 317) having a first diameter and emitting light having a first wavelength. A detector (118, 218, 500) is positioned to receive light emitted from the plurality of light emitting particles (108, 208, 317) and a processing device (500) determines the pulse rate. The light emitting device (100, 201, 310, 500) and the detector (118, 218, 500) are disposed on a flexible polymeric material (102, 202, 334). The light emitting device (100, 201, 310, 500) may include a second plurality of light emitting particles (108, 208, 317) having a second diameter and emitting light having a second wavelength, wherein the processing device (500) determines the user's blood oxygen level. The light emitting particles (108, 208, 317) may comprise one of quantum dots, electroluminescent particles, or organic particles.