Samsung Electronics Co., Ltd.

Korea Electronics Technology Institute

Hyundai Electronics Industries Co. Ltd.

Korea University of Technology and Education Industry-University Cooperation Foundation

Young Bo Shim

Kyung Shik Roh

Young Do Kwon

Jeong Hun Kim

Seon Young Lee

Yong Jae Kim

Suk June Yoon

San Lim

Youn Baek Lee

Dong Soo Kim

Ji Young Kim

Sung Hwan Ahn

Jong Won Lee

Hyun Do Choi

Se-Gon Roh

Youngjin Park

Yong-Jae Kim

Hyo Seok Hwang

Woong Kwon

Bokman Lim

Seungyong Hyung

Won Jun Hwang

Tae Sin Ha

Ji Min Kim

Min Hyung Lee

Soo Sang Yang

Seung Yong Hyung

Jeong Hwan Kim

Joon-Kee Cho

Byung-Kwon Choi

Minho Choi

Myung Hee Kim

Sunghwan Ahn

Haeng Seon Son

Kyoung Won Min

Jong Do Choi

Byung June Choi

Yeon Kug Moon

Chang Gue Park

Yeon Taek Oh

Ja Woo Lee

Kwang Jun Kim

Gyung Rock Kim

Hyun Seok Yang

Ganzorig Gankhuyag

Hyun Kyu Kim

Kee Hong Seo

No San Kwak

Bok Man Lim

Chang Hyun Roh

Ju Suk Lee

Ji Hyo Lee

Se Gon Roh

JeongHun Kim

Xiangsheng Huang

Haewook Ahn

Woo Seok Han

Kyung-Won Moon

JongWon Lee

Mikyung Park

Dong Jun Hyun

Jeong-heon Han

Jeong Heon Han

Gi Ho Sung

Yeong Kwon Choe

Jin Man Park

Huang Xiangsheng

Yeji Bae

Jong Kyeong Shin

There are provided a scenario similarity retrieval-based automatic scenario generation system and method. According to an embodiment, a scenario retrieval-based automatic scenario generation method includes: retrieving scenarios similar to a query scenario from a scenario DB for an autonomous driving test; filtering only scenarios that meet a selection condition from the retrieved scenarios; and converting components of the filtered scenarios to suit a target condition. Accordingly, a desired scenario may be automatically generated by retrieving a scenario similar to a targeted scenario, converting the retrieved scenario, and concretizing the scenario.

Scenario similarity retrieval-based automatic scenario generation system and method

There are provided a behavior recognition method for efficient recognition of hand signals and gesture and a behavior recognition AI network system using the same. The behavior recognition method for efficient recognition of hand signals and gestures according to an embodiment includes: an input feature extraction step of extracting, by a behavior recognition AI network system, key points Ffrom bounding box data of an object which makes hand signals to be inputted by sequence, and generating skeleton data of the object; and a spatial feature extraction step of calculating, by the behavior recognition AI network system, a length Fand an angle Fof a bone vector based on the key points Fand the skeleton data, and extracting a spatial feature. Accordingly, performance of recognition of hand signals and gestures may be enhanced.

Behavior recognition artificial intelligence network system and method for efficient recognition of hand signals and gestures

Provided is an air housing apparatus for protecting, from contamination, scratches, damage and the like, a lens and a cover glass which affect the performance of an optical device installed on a vehicle such as a camera and LIDAR. The air housing apparatus according to an embodiment of the present invention comprises: a cover into which high pressure air is fed; an air guide for injecting the supplied high pressure air toward the front surface of a lens or a sensor; and an air housing provided so as to connect the cover and the air guide and transferring the high pressure air supplied from the cover to the air guide. Therefore, a lens, sensor, or cover glass of a vehicle-installed optical device such as a camera and LIDAR can be continually protected from contamination and scratches by means of the injected air. In addition, energy efficiency can be enhanced by controlling the discharge pressure of air provided by an air pressure generation apparatus on the basis of information relating to the running speed and RPM of the vehicle or the air flow direction and air flow velocity.

Air housing apparatus for protecting lens of vehicle-installed optical device

Provided is a method for filtering driving paths in order to stably track a path which is generated for various purposes, such as recognizing a lane and keeping or changing the lane, or avoiding an obstacle around a road. A driving path filtering system according to an embodiment of the present disclosure includes: a position recognition unit configured to recognize a position of a vehicle; a lane recognition unit configured to recognize a lane of a road; and a path generator configured to generate a path for the vehicle to travel on, based on a result of recognizing the position and a result of recognizing the lane, and to perform path noise filtering with respect to the generated path. Accordingly, a driving path can be stabilized by removing a path noise (error) by complementally filtering a lane recognition-based path generation method and a map-based path generation method according to a driving environment.

Method for filtering the driving route based on distributed Kalman filter for stabilization of route tracking of autonomous vehicles

Provided is a deep learning-based panoptic segmentation accelerated processing technique using a complexity-based RPN skip method. An image segmentation system includes: a first processing unit configured to extract dynamic objects in an instance segmentation method by using an extracted feature; a calculation unit configured to control to skip some areas of the feature extracted at the network by the first processing unit, on the basis of complexity of the input image; a second processing unit configured to extract static objects in a semantic segmentation method by using the feature extracted at the network; and a fusion unit configured to fuse a result of extracting by the first processing unit and a result of extracting by the second processing unit. Accordingly, the panoptic segmentation method can be easily performed even in an embedded environment by reducing complexity for panoptic segmentation processing by reducing a calculation burden.

Accelerated processing method for deep learning based-panoptic segmentation using a RPN skip based on complexity

The described technology relates to a light detection and ranging (LIDAR) device. The LIDAR device can include a transmission unit configured to emit a first signal, a first lens unit configured to convert the first signal into parallel light, a reflection unit configured to adjust a direction of the converted first signal and a second lens unit configured to enable the first signal to have the same focal plane even when a reflection angle of the reflection unit changes. The LIDAR device can also include a third lens unit configured to convert the first signal passing through the second lens unit into parallel light, a fourth lens unit configured to increase an angle of the first signal passing through the third lens unit and a reception unit configured to receive a second signal reflected by an object after passing through the fourth lens unit.

LIDAR device and LIDAR system including the same

The described technology relates to a light detection and ranging (LIDAR) device. The LIDAR device can include a transmitter configured to emit an optical signal, a first lens section configured to convert the optical signal into collimated light, a reflector configured to adjust a direction of the converted optical signal, a second lens section configured to allow the adjusted optical signal to have the same focal plane even though a reflection angle of the reflector is varied and a third lens section configured to convert the optical signal passed through the second lens section into collimated light. The LIDAR device can also include a fourth lens section configured to allow the optical signal, and a receiver configured to receive the optical signal passed through the fourth lens section. The third lens section and the fourth lens section are positioned on the same line in a first direction.

LIDAR device and system comprising the same

A light detection and ranging (LIDAR) time of flight (TOF) sensor for inputting and outputting simultaneously and 3-dimensional laser scanning system including the same are disclosed. In one aspect, the sensor includes a substrate and a light receiving element array provided on the substrate and including a plurality of light receiving elements. The sensor also includes readout circuits configured to receive electrical signals from the light receiving elements and perform signal processing on the electrical signals. The sensor further includes metal lines disposed on the light receiving element array in parallel, provided to correspond to the number of the light receiving elements, and configured to connect the light receiving elements to the readout circuits in one-to-one correspondence.

Light detection and ranging (LIDAR) time of flight (TOF) sensor capable of inputting and outputting simultaneously and 3-dimensional laser scanning system including the same

A motion assistive apparatus may include a receiver allowing a user to intuitively adjust performance of the motion assistive apparatus and a method of controlling the same. The motion assistive apparatus may include a receiver provided to adjust variable characteristics of the motion assistive apparatus, a processor to adjust a variable parameter related to adjustment of the variable characteristics, in response to adjustment of the variable characteristics through the receiver, and an actuator to output changed assist power in response to adjustment of the variable parameter to change the variable characteristics.

Motion assistive apparatus and method of controlling the same

Disclosed herein is a control method of a wearable robot, including: generating reference gait data based on the results of sensing by a sensor unit included in a structure; estimating, when a wearer walks, the wearer's gait phase based on the results of sensing by the sensor unit; detecting a gait phase having a minimum difference from the estimated gait phase from the reference gait data; and driving a driver of the structure, according to a control signal generated based on the estimated gait phase and the detected gait phase.

Wearable robot and control method thereof

There is provided a method of controlling a wearable robot. The method includes measuring an electrical signal from a scalp of a wearer, estimating a current walking speed of the wearer using the measured electrical signal, and outputting assistive torque which allows the estimated current walking speed to approximate a target walking speed.

Wearable robot and method of controlling the same

A walk-assistive apparatus may include at least one joint that corresponds to at least one joint of a wearer, at least one link that connects the joint, and is rotated in response to rotation of the joint, a spring that is mounted in the link or the joint so that a length of the spring is changed in accordance with rotation of the link or the joint, and a processor that controls the change in the length of the spring to compensate for a weight by gravity when the wearer walks. Accordingly, the walk-assistive apparatus and a method of controlling the walk-assistive apparatus may use a mechanical element such as a spring to reduce energy, and weight compensation having uniform performance may be performed even in an arbitrary posture.

Walk-assistive apparatus and method of controlling the walk-assistive apparatus

Provided are a walk-assistive robot and a method of controlling the same. The method of controlling the walk-assistive robot includes: obtaining ground information that is information regarding ground a walking direction; determining control patterns of the walk-assistive robot by analyzing the obtained ground information; and controlling the walk-assistive robot based on the determined control patterns.

Walk-assistive robot and method of controlling the same

Provided is a method of controlling a wearable robot, the method including: measuring a ground reaction force (GRF) exerted on a wearer's soles; calculating a time variation rate of the measured GRF; measuring the wearer's knee joint angle; and detecting a time point at which the calculated time variation rate of the GRF and the measured knee joint angle cross each other.

A walking assistance method may include: computing an amount of exercise of a user based on a biosignal of the user; adjusting a pattern of an assist parameter based on the amount of exercise; and/or generating a force corresponding to the amount of exercise, based on the adjusted pattern. A walking assistance apparatus may include: a pattern adjuster configured to compute an amount of exercise of a user based on a biosignal of the user; and/or a driver configured to generate a force corresponding to the amount of exercise based on a pattern of an assist parameter based on the amount of exercise.

Walking assistance methods and apparatuses performing the same

Growing community of inventors

Seoul, South Korea

Young Bo Shim