Abstract: There is provided a training method of a multi-task integrated deep learning model. A multi-task integrated deep learning model training method according to an embodiment may generate training data for a plurality of visual intelligence tasks from visual data in a batch, and may train a multi-task integrated deep learning model which performs a plurality of visual intelligence tasks by using the generated training data. Accordingly, training data for training an integrated deep learning model which performs various visual intelligence tasks is generated in a batch through multi-data conversion kernels, so that appropriate training data for performing multiple tasks may be easily obtained and effective training of a multi-task integrated deep learning model is possible.

Abstract: Provided are a laser element and a method for manufacturing the same, in which the laser element includes a first clad layer, an optical waveguide disposed on the first clad layer, a second clad layer disposed on the optical waveguide, a first electrode disposed on the second clad layer, and a dummy clad disposed on the optical waveguide and apart from the second clad layer and the first electrode.

Abstract: There are provided an apparatus and a method for reconstructing a 3D human object in real time based on a monocular color image. A 3D human object reconstruction apparatus according to an embodiment extracts a pixel-aligned feature from a monocular image, extracts a ray-invariant feature from the pixel-aligned feature, generates encoded position information by encoding position information of a point, predicts a SD of a point from the ray-invariant feature and the encoded position information which are extracted, and reconstructs a 3D human object by using the predicted SD. Accordingly, the ray-invariant feature extracted from the pixel-aligned feature, and the encoded position information are used, so that an amount of computation for predicting SDs of points of a 3D space can be noticeably reduced and a speed can be remarkably enhanced.

Abstract: Provided is an illumination device including a light extraction film having a light guide plate including one or more edges and including a transparent substrate and a light extraction film on the light guide plate, where the film includes a matrix layer, and a plurality of scattering particles embedded in the matrix layer, and the plate includes first areas in which a volume density of the particles of the light extraction film is substantially constant, and second areas in which a volume density of the particles of the light extraction film varies along a Gaussian profile.

Abstract: A method for providing a cloud-based application programming interface (API) service is provided. The method may include issuing an authentication key corresponding to an application for some of a plurality of API services, performing authentication for a first user using the issued authentication key in response to a call for a first API service by the first user, and executing the first API service based on a result of the authentication.

Abstract: There is provided a cloud-based application programming interface (API) service providing system. The system may comprise a cloud management console issuing authentication information for a requested API among a plurality of APIs, an API manager receiving a call message for a specific API, which comprises authentication information, from a user's terminal, performing authentication using the authentication information in the call message, and requesting the execution of the specific API based on the authentication result and an API execution engine generating the result of executing the specific API in response to a request from the API manager, wherein the cloud management console is a module belonging to a control plane, and the call message is received by the API manager belonging to a data plane without via the control plane.

Abstract: There is provided a learning method and system of a backbone network for visual intelligence based on self-supervised learning and multi-head. A network learning system according to an embodiment generates a plurality of first modified vectors by modifying a first feature vector outputted from a teacher network, generates a plurality of second modified vectors by modifying a second feature vector outputted from a student network, calculates a loss by using the first modified vectors and the second modified vectors, and optimizes parameters of the student network. Accordingly, the effect of learning by knowledge distillation may be enhanced by training the backbone network for visual intelligence like group learning is performed by various teacher networks and student networks.

Abstract: There are provided an apparatus and a method for reconstructing a 3D human object based on a monocular image through depth image-based implicit function learning. A 3D human object reconstruction method according to an embodiment includes: predicting a double-sided orthographic depth map from a front perspective color image of a human object; predicting a signed distance (SD) regarding points on a 3D space from the predicted double-sided orthographic depth map; and reconstructing a 3D human object by using the predicted SD. Accordingly, a human object and details can be naturally reconstructed with respect to not only an area visible through a front perspective color image of the human object but also an invisible area.

Abstract: There is provided an emotion prediction method based on virtual facial expression image augmentation. The emotion prediction method may acquire a user facial image, may extract a facial expression feature from the acquired user facial image, and may predict a user emotion from the extracted facial expression feature. The emotion prediction method may extract the facial expression feature by using a facial expression recognition network, the facial expression recognition network being an AI model that is trained to receive a user facial image and to extract a facial expression feature. The facial expression recognition network is retrained with virtual facial images which are augmented from a facial image that causes a failure in emotion recognition. Accordingly, by augmenting features of a facial expression image that causes a failure in prediction through error feedback, facial expression recognition performance can be enhanced.

Abstract: There is provided a depth estimation method for a small baseline-stereo camera through LiDAR sensor fusion. A depth map estimation method according to an embodiment may estimate a high-resolution depth map from a small baseline-stereo image based on deep learning, by using transfer learning from a deep learning network that is trained to estimate a depth map from a wide baseline-stereo image. Accordingly, in a device which has a small baseline-stereo camera installed therein due to structural constraints, such as a smartphone, a wearable AR/VR device, a drone, 3D image quality can be enhanced. In addition, according to embodiments, pseudo-LiDAR data may be generated by using a depth map estimated from a small baseline-stereo image, and may be used for replacing or reinforcing LiDAR data.

Abstract: According to embodiments, modeling through evaluation of thermal-hydraulic analysis of a unique containment facility may be provided. Also, orders for nuclear power plants in other countries may be obtained, and a relative technological advantage may be achieved among nuclear power plant developing countries. In addition, a power plant may be designed through conservative and realistic modeling, and costs required for power plant design may be reduced by optimizing a design margin of a containment facility.

Abstract: A display area (100a) includes a plurality of tiles each tile including an optically transparent layer (208). Each of the plurality of tiles includes a substrate (202) and a plurality of light sources (204?) positioned on the substrate. In addition, each of the plurality of tiles includes a non-transparent layer (206) extending between the transparent layer and the plurality of light sources. The non-transparent layer includes an opening (214) extending over an area between at least one of the plurality of light sources and the transparent layer. In addition, a seam extends between adjacent tiles of the plurality of tiles and the non-transparent layer extends across the seam.

Abstract: There is provided a self-directed visual intelligence system, The self-directed visual intelligence system according to an embodiment prepares data necessary for training a visual intelligence model when a change in a visual context of a real world is recognized, configures a visual intelligence model and configures training data of the visual intelligence model, based on the changed visual context of the real world, trains the configured visual intelligence model with the training data, and evaluates performance of the trained visual intelligence model. Accordingly, the visual intelligence model is corrected/improved in a self-directed way according to a change in a visual context of a real world, and is grown/advanced by itself, so that performance of the visual intelligence model is maintained in a best state even in response to any change in the context of the real world.

Abstract: A method and an apparatus for detecting a lane is provided. The lane detection apparatus according to an embodiment includes: an acquisition unit configured to acquire a front image of a vehicle; and a processor configured to input the image acquired through the acquisition unit to an AI model, and to detect information of a lane on a road, and the AI model is trained to detect lane information that is expressed in a plane form from an input image. Accordingly, data imbalance between a lane area and a non-lane area can be solved by using the AI model which learns/predicts lane information that is expressed in a plane form, not in a segment form such as a straight line or curved line.

Abstract: An embodiment electrical connection device for a vehicle includes a male connector housing having a blade terminal mounting portion, a female connector housing to couple to the male connector housing and having a tab terminal mounting portion, a blade terminal mounted in the blade terminal mounting portion, and a tab terminal mounted in the tab terminal mounting portion to connect to the blade terminal when the male connector housing and the female connector housing are coupled to each other. An embodiment blade terminal includes a terminal body having a first connection portion and a contact spring inserted into an inner side of the first connection portion to connect the terminal body and the tab terminal to each other by being brought into contact with the tab terminal when the tab terminal is inserted into the inner side of the first connection portion.

Abstract: Various aspects of a light extraction substrate, an organic light emitting device, and methods of fabrication are provided. A light extraction substrate of an organic light-emitting device includes a light-scattering layer disposed on a base substrate and contains a first material, and a number of holes (hole diameters ranging from 350 nm to 450 nm) extending between the first surface and the second surface. A planarization layer (thickness not greater than 200 nm) is disposed on the light-scattering layer and contains a second material.

Abstract: Provided is a shelf lighting device. A light guide plate is disposed below a shelf to which external contacts are made, includes a first surface facing the shelf, a second surface opposite the first surface, and third surfaces extending between edges of the first and second surfaces to connect the first and second surfaces, is formed from a first refractive index material, and guides light. An area in which light is guided is separated from an area in which external contacts are made, such that light is not extracted directly from a stained portion, thereby reducing sensitivity to surface contamination.

Abstract: According to an embodiment, a wavelength tunable laser comprising a gain region and a wavelength tunable area is disclosed. The wavelength tunable area comprises: a lower clad layer; a passive optical waveguide positioned on the lower clad layer; an upper clad layer positioned on the passive optical waveguide; a drive electrode positioned on the upper clad layer; a current blocking layer positioned on the drive electrode; a heater positioned on the current blocking layer; and a first insulating groove and a second insulating groove which are positioned so as to face each other with the passive optical waveguide therebetween.

Abstract: Provided are an integrated light guide plate and a lighting device including the same. Micro-sized engraved lenses are provided on one surface of a transparent substrate, coated with a reflector pattern having a visually unrecognizable size, such that dual-surface lighting is realized and the ratio between intensities of dual-surface lighting is adjusted without the addition of a reflector plate and a diffuser plate. The transparent substrate includes a first surface, a second surface, and third surfaces connecting the first surface and the second surface to each other, the second surface provided with a lens pattern including engraved lenses.

Abstract: An optical display device having a backplane substrate and a cover plate adjacent to and spaced apart from the backplane substrate is provided. The backplane substrate can include a plurality of electroluminescent elements disposed thereon and the cover plate can include a plurality of light absorbing wedge-shaped features arranged in rows thereon.