Dong-Hyun Hwang

Dong-Hyun Hwang
Naver

Doctor of Philosophy
ML Researcher@CLOVA Voice&Avatar, NAVER

About

21
Publications
6,992
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
84
Citations
Introduction
I am a researcher at CLOVA Avatar, NAVER Corp. I got a bachelor's degree in Science in Computer Engineering from Korea University of Technology and Education and master's and Ph.D. degrees in computer science from Tokyo Institute of Technology.
Additional affiliations
April 2019 - March 2022
Tokyo Institute of Technology
Position
  • PhD Student
April 2017 - March 2019
Tokyo Institute of Technology
Position
  • Master's Student
Education
April 2017 - March 2022
Tokyo Institute of Technology
Field of study
  • Computer Science
March 2010 - February 2017
Korea University of Technology and Education
Field of study
  • Computer Science and Engineering

Publications

Publications (21)
Article
Full-text available
In this paper, We proposed a novel hand-gesture recognition algorithm using concentric-circle expanding and tracing. The proposed algorithm determines region of interest of hand image through preprocessing the original image acquired by web-camera and extracts the feature of hand gesture such as the number of stretched fingers, finger tips and fing...
Conference Paper
Full-text available
In this work, we propose a simple yet effective method for synthesizing a pseudo-2.5D scene from a monocular video for mixed reality (MR) content. We also propose the ParaPara system, which applies this method. Most previously proposed systems convert real-world objects into 3D graphic models using expensive equipment; this is a barrier for individ...
Conference Paper
Full-text available
We propose MlioLight, a projector-camera unit-based projection-mapping system for overlaying multiple images on a screen or on real world objects using multiple flashlight-type devices. We focus on detecting the areas of overlapping lights in a multiple light source scenario and overlaying multi-layered information on real world objects in these ar...
Conference Paper
Full-text available
We present MoVNect, a lightweight deep neural network to capture 3D human pose using a single RGB camera. To improve the overall performance of the model, we apply the teacher-student learning method based knowledge distillation to 3D human pose estimation. Real-time postprocessing makes the CNN output yield temporally stable 3D skeletal informatio...
Conference Paper
Full-text available
We present MonoEye, a multimodal human motion capture system using a single RGB camera with an ultra-wide fisheye lens, mounted on the user's chest. Existing optical motion capture systems use multiple cameras, which are synchronized and require camera calibration. These systems also have usability constraints that limit the user's movement and ope...
Article
Full-text available
MonoMR is a system that synthesizes pseudo-2.5D content from monocular videos for mixed reality (MR) head-mounted displays (HMDs). Unlike conventional systems that require multiple cameras, the MonoMR system can be used by casual end-users to generate MR content from a single camera only. In order to synthesize the content, the system detects peopl...
Preprint
Full-text available
In this work, we present an analysis tool to help golf beginners compare their swing motion with experts' swing motion. The proposed application synchronizes videos with different swing phase timings using the latent features extracted by a neural network-based encoder and detects key frames where discrepant motions occur. We visualize synchronized...
Patent
A motion measurement system including a wide-angle camera configured to capture in the periphery of an image at least a part of a body of a subject when the wide-angle camera is mounted on the body, a feature point extractor configured to extract feature points from the image, and a 3D pose estimator configured to estimate 3D pose data of the subje...
Patent
Provided is a method of generating data for estimating a three-dimensional (3D) pose of an object included in an input image, the method including acquiring the input image including at least one moving object, estimating location information about each joint of a plurality of joints of the object included in the input image using a prediction mode...
Preprint
Full-text available
We present MoVNect, a lightweight deep neural network to capture 3D human pose using a single RGB camera. To improve the overall performance of the model, we apply the teacher-student learning method based knowledge distillation to 3D human pose estimation. A real-time post-processing makes the CNN output to yield temporally stable 3D skeletal info...
Conference Paper
This paper proposed a golf training system using real-time audio-visual feedback. The system captures user's motion with an optical motion tracking system and projects his/her posture as a virtual shadow on the ground. Unlike other golf training systems, our system enables the user to keep his/her gaze on the ball. The model swing motion of the exp...
Conference Paper
Full-text available
Wearable cameras have the potential to be used in various ways in combination with egocentric views such as action recognition, gesture input method for augmented/virtual reality (AR/VR) as well as lifelogger. Particularly, the pose of the camera wearer is one of the interesting factors of the egocentric view and various eccentric view-based pose e...
Conference Paper
Full-text available
In this work, we propose a golf training system using real-time visual feedback. The system projects the virtual shadow of the user on the ground in front of the user; this shadow provides feedback to the user without form collapse. Additionally, an expert's contour is overlaid on the virtual shadow of the user to make them aware of the difference...
Conference Paper
Full-text available
A self-sports learning system that provides users with real-time multimodal feedback about differences between a user's motion and an expert's motion is proposed. We also propose the Decayed Dynamic Time Warping algorithm, which allows the user to change the motion speed dynamically and repeat a target motion without additional operations. The user...
Article
In this paper, we propose a novel algorithm, Concentric-Circle Tracing algorithm, which recognizes finger's shape and counts the number of fingers of hand using low-cost web-camera. We improve algorithm’s usability by using low-price web-camera and also enhance user's comfortability by not using a additional marker or sensor. As well as counting th...

Network

Cited By

Projects

Projects (3)
Archived project
In this system, we propose novel information passing system for large size table-top display.
Archived project
We compare existing raster scan algorithm and our concentric-circle algorithm when it's applied hand gesture recognition.