Home
대학원생 및 연구원 모집
We study mathematics and machine learning-based image processing with various applications in media, culture, and industries. Our current research topics include 3D volumetric geometric coding for ultra-realistic light-field media, restoration of Korean national heritage wooden buildings, and 3D surface geometric data representation. We have strong ties to local industries. We developed and successfully deployed AI solutions for steelmaking, shipbuilding, and automotive manufacturing industries.
SPL seeks graduate students and researchers interested in mathematics and machine learning-based image processing. Current research posts are related to point cloud data representation, recognition, compression, and generation. If you want to work in our lab, please email Professor Seungjoon Yang.
- SPL 연구실에서 대학원생 모집합니다.
전기전자공학과대학원 (oo명) 인공지능대학원 (oo명)
- SPL 연구실에서 박사후 연구원 모집합니다.
인공지능대학원 InnoCore 사업 (oo명)
Research
News
- All
- Projects
- Publication
- Research
Development of AI Agent Technology for Architectural Heritage Restoration Design
국립문화유산연구원 2026-04 ~ 2029-12
TSDF Volume Compression using Sampling-Enhancing Residual Block and Selective Latent Code Encoding, ACM TOMM 2026
SLTN: Shadow and Lighting Transformation Network for Efficient 3D Shape Recognition, ICASSP 2026
We propose SLTN, a Shadow and Lighting Transformation Network for efficient 3D shape recognition. Unlike prior methods that only optimize viewpoints, SLTN jointly learns camera poses and illumination, leveraging a differentiable shadow renderer to integrate cast shadows as discriminative cues. This joint optimization reduces viewpoint...
Generalization of Real World Video Deblurring By Image-to-Image Translation, WACV 2026
We address the challenge of generalizing video deblurring models to real-world scenarios, where traditional methods often fail due to a significant domain gap between synthetic and real blur. This work extends the image-to-image translation framework to the more complex domain of video deblurring, introducing a...
TSDF-Based Efficient Motion-Compensated Temporal Interpolation for 3D Dynamic Sequences, AAAI 2025
This paper introduces a method for efficiently interpolating 3D dynamic sequences using truncated signed distance function (TSDF) volumes. The method calculates bi-directional motions between TSDF volumes of two frames and refines them to reconstruct intermediate frames. Unlike point cloud-based methods, which can suffer from varying...
AI-based traditional wooden architecture analysis and design
National Research Institute of Cultural Heritage (2024~2025)
Nakkwan Choi, et.al. Restoration of Hand-Drawn Architectural Drawings using Latent Space Mapping with Degradation Generator. CVPR 2023
This work presents the restoration of drawings of wooden built heritage. Hand-drawn drawings contain the most important original information but are often severely degraded over time. A novel restoration method based on the vector quantized variational autoencoders is presented. Latent space representations of drawings and...
최낙관, et al. 이미지 생성 및 지도학습을 통한 전통 건축 도면 노이즈 제거. 건축역사연구, 2022, 31.1: 41-50.
Traditional wooden buildings deform over time and are vulnerable to fire or earthquakes. Therefore, traditional wooden buildings require continuous management and repair, and securing architectural drawings is essential for repair and restoration. Unlike modernized CAD drawings, traditional wooden building drawings scan and store hand-drawn drawings,...
CHO, Hyunjoong, et al. Online Safety Zone Estimation and Violation Detection for Nonstationary Objects in Workplaces. IEEE Access, 2022, 10: 39769-39781.
This study presents a deep neural network (DNN)-based safety monitoring method. Nonstationary objects such as moving workers, heavy equipment, and pallets were detected, and their trajectories were tracked. Time-varying safety zones (SZs) of moving objects were estimated based on their trajectories, velocities, proceeding directions, and...
LEE, Yongsik; JANG, Jinhyuk; YANG, Seungjoon. Virtual portraits from rotating selfies. ETRI Journal, 2022.
Selfies are a popular form of photography. However, due to physical constraints, the compositions of selfies are limited. We present algorithms for creating virtual portraits with interesting compositions from a set of selfies. The selfies are taken at the same location while the user spins...
Ship production progress monitoring base on image and CAD data recognition
Korea Shipbuilding & Offshore Engineering Co., Ltd. KSOE 2022
Ship production progress monitoring base on humanless image acquisition system
Korea Shipbuilding & Offshore Engineering Co., Ltd. KSOE 2022
Modeling and restoration of image degradation
Related Projects: AI-based traditional architectural hand drawing to CAD conversion Nonstationary and asymmetric lens blur restoration 8K 120Hz AR/VR display SoC HDR color image processing Embedded image processing for low-power LED public displays Smart pico projector parts and engine for 3D HD images Layerwise super...
Light field media coding and its applications
Related Project: Audio/video coding and light field media fundamental technologies for ultra realistic tera-media Object detection and distortion correction for 360 view camera system Computational photography with rotating selfies Multi-view image acquisition system for face recognition Dynamic view generation for augmented reality OpenGL based SMMD...
Optimal architecture for deep neural networks
Related Project: UniBrain – Ultimate neuromorphic intelligent brain system engineering
Industrial and vehicular safety
Related Project: Self-elevating crane system for installation and maintenance of onshore wind power systems AI-based simultaneous product classification and quality evaluation Abnormal behavior detection in loading dock Inteligent image masking Pedestrian behavior recognition Back seat passenger detection Personal protective equipment detection and safety distance monitoring...
Cho, H., Jang, J., Lee, C. and Yang, S., 2021. Efficient architecture for deep neural networks with heterogeneous sensitivity. Neural Networks, 134, pp.95-106.
Abstract - This work presents a neural network that consists of nodes with heterogeneous sensitivity. Each node in a network is assigned a variable that determines the sensitivity with which it learns to perform a given task. The network is trained by a constrained optimization...
Self-elevating crane system for installation and maintenance of onshore wind power systems
Korea Energy Technology Evaluation and Planning 2021-2024
Audio/video coding and light field media fundamental technologies for ultra realistic tera-media
ETRI 2021-2022
AI-based simultaneous product classification and quality evaluation
Hanguk mold Co., Ltd./UNIST 2021
Gwak, M. and Yang, S., 2020. Modeling nonstationary lens blur using eigen blur kernels for restoration. Optics Express, 28(26), pp.39501-39523.
Abstract - Images acquired through a lens show nonstationary blur due to defocus and optical aberrations. This paper presents a method for accurately modeling nonstationary lens blur using eigen blur kernels obtained from samples of blur kernels through principal component analysis. Pixelwise variant nonstationary lens blur...
Jang, J., Cho, H., Kim, J., Lee, J. and Yang, S., 2020. Deep neural networks with a set of node-wise varying activation functions. Neural Networks, 126, pp.118-131.
Abstract - In this study, we present deep neural networks with a set of node-wise varying activation functions. The feature-learning abilities of the nodes are affected by the selected activation functions, where the nodes with smaller indices become increasingly more sensitive during training. As a result,...
8K 120Hz AR/VR display SoC
Korea Evaluation Institute of Industrial Technology, KEIT 2020-2022
Gwak, M., Lee, C., Lee, H., Cheong, W.S. and Yang, S., 2019. Moving Object Preserving Seamline Estimation. Journal of Broadcast Engineering, 24(6), pp.992-1001.
Abstract - In many applications, images acquired from multiple cameras are stitched to form an image with a wide viewing angle. We propose a method of estimating a seam line using motion information to stitch multiple images without distortion of the moving object. Existing seam estimation...
Park, E., Han, B.J., Yang, S. and Sim, J.Y., 2018, November. Video Saliency Detection Using Adaptive Feature Combination and Localized Saliency Computation. In 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) (pp. 498-504). IEEE.
Abstract - A novel saliency detection algorithm for videos is proposed in this paper. We adaptively determine the weights of color and motion features to extract combined global feature contrast by adopting the compactness prior of salient object. We localize a saliency searching area in a...
Cho, H., Baek, Y.S., Kwak, Y. and Yang, S., 2018. Estimation of Perceptual Surface Property Using Deep Networks With Attention Models. IEEE Access, 6, pp.72173-72178.
Abstract - How we perceive property of surfaces with distinct geometry and reflectance under various illumination conditions is not fully understood. One widely studied approach to understanding perceptual surface property is to derive statistics from images of surfaces with the goal of constructing models that can...
Kim, S., Gwak, M. and Yang, S., 2018. Non-stationary deep network for restoration of non-Stationary lens blur. Pattern Recognition Letters, 108, pp.62-69.
Abstract - Optical aberrations of a lens introduce lens blur to photographed images. Lens blur is non-stationary with the amount and characteristics of blur varying depending on spatial pixel locations in an image. This work presents non-stationary deep networks for the restoration of non-stationary lens blur. Deep networks...
Jang, J., Cho, H., Kim, J., Lee, J. and Yang, S., 2018. Facial attribute recognition by recurrent learning with visual fixation. IEEE transactions on cybernetics, 49(2), pp.616-625.
Abstract - This paper presents a recurrent learning-based facial attribute recognition method that mimics human observers' visual fixation. The concentrated views of a human observer while focusing and exploring parts of a facial image over time are generated and fed into a recurrent network. The network...
Infra-red camera based vehicular part quality evaluation
Korea Evaluation Institute of Industrial Technology, KEIT 2018
Lee, Y. and Yang, S., 2017. Parallel block sequential closed-form matting with fan-shaped partitions. IEEE Transactions on Image Processing, 27(2), pp.594-605.
Abstract - Applying alpha matting to large images is a challenging task because of its computational complexity. This paper provides a divide and conquer strategy for performing closed-form matting. The matting problem, defined for an entire image, is broken down into systems of linear equations defined...
Yun, J.D. and Yang, S., 2017. Admm in krylov subspace and its application to total variation restoration of spatially variant blur. SIAM Journal on Imaging Sciences, 10(2), pp.484-507.
Abstract - In this paper we propose an efficient method for a convex optimization problem which involves a large nonsymmetric and non-Toeplitz matrix. The proposed method is an instantiation of the alternating direction method of multipliers applied in Krylov subspace. Our method offers significant advantages in...
Nonstationary and asymmetric lens blur restoration
National Research Foundation of Korea, NRF 2016~2022
Kim, S., Park, B., Song, B.S. and Yang, S., 2016. Deep belief network based statistical feature learning for fingerprint liveness detection. Pattern Recognition Letters, 77, pp.58-65.
Abstract - Fingerprint recognition systems are vulnerable to impersonation by fake or spoof fingerprints. Fingerprint liveness detection is a step to ensure whether a scanned fingerprint is live or fake prior to a recognition step. This paper presents a fingerprint liveness detection method based on a...
Jang, J., Yun, J.D. and Yang, S., 2016. Modeling non-stationary asymmetric lens blur by normal sinh-arcsinh model. IEEE Transactions on Image Processing, 25(5), pp.2184-2195.
Abstract - Images acquired by a camera show lens blur due to imperfection in the optical system even when images are properly focused. Lens blur is non-stationary in a sense that the amount of blur depends on pixel locations in a sensor. Lens blur is also...
Kim, S.H., Park, R.H., Yang, S. and Kim, H.Y., 2016. Feature-Based Affine Motion Estimation for Superresolution of a Region of Interest. In Human-Computer Interaction: Concepts, Methodologies, Tools, and Applications (pp. 682-701). IGI Global.
Abstract - This chapter presents an interpolation method of low-computation for a Region Of Interest (ROI) using multiple low-resolution images of the same scene. Interpolation methods using multiple images require the accurate motion information between the reference image of interpolation and the other images. Sometimes...
Lee, H.G., Yang, S. and Sim, J.Y., 2015, December. Color preserving contrast enhancement for low light level images based on Retinex. In 2015 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA) (pp. 884-887). IEEE.
Abstract - In this paper, we propose a novel contrast enhancement algorithm for low light level images, which preserves image details and color constancy based on Retinex. We decompose an input low contrast image into luminance and chrominance components in Lab color space, which reflects the...
Kim, S. and Yang, S., 2015, September. Environmental illumination invariant face recognition using near infrared imaging system. In 2015 9th International Symposium on Image and Signal Processing and Analysis (ISPA) (pp. 87-92). IEEE.
Abstract - Environmental illumination can degrade performance of face recognition systems. This paper presents a method to reduce the effect of environmental illumination for face recognition with near infrared (NIR) imaging systems. The camera response function that maps the image irradiance to image brightness is estimated...
Baek, Y.S., Kwak, Y. and Yang, S., 2015. Visual appearance measurement of surfaces containing pearl flakes. JOSA A, 32(5), pp.934-942.
Abstract - The color, gloss, and texture (i.e., pearliness) of 15 glossy samples containing pearl flakes were investigated. Psychophysical experimental data from 21 observers were compared with measurement data. Color measurement data obtained using the CIE D/0 and ASTM E2539-08 multiangle geometry did not predict the...
Yun, J.D., Park, J.H. and Yang, S., 2015. Efficient nonlinear geometric distortion correction for scanned laser pico projectors. Electronics Letters, 51(6), pp.471-473.
Abstract - Scanned laser pico projectors utilising laser light sources and microelectromechanical systems mirror devices introduce geometric distortion to the displayed images because of nonlinear mirror driving methods. A pre-processing algorithm is proposed that corrects the nonlinear geometric distortion of scanned laser pico projectors. The distortion...
Smart pico projector parts and engine for 3D HD images
Korea Evaluation Institute of Industrial Technology, KEIT 2013~2016
Yun, J.D., Kwak, Y. and Yang, S., 2013. Evaluation of perceptual resolution and crosstalk in stereoscopic displays. Journal of Display Technology, 9(2), pp.106-111.
Abstract - Perceptual resolution and crosstalk of active and passive stereoscopic 3D displays are evaluated by subjective tests. Test patterns of gratings with various contrast and spatial frequencies are used to determine the contrast thresholds, from which the contrast sensitivity of human visual systems is obtained....
Kwak, Y., Lee, S. and Yang, S., 2012. Crosstalk characterization method for stereoscopic three-dimensional television. IEEE Transactions on Consumer Electronics, 58(4), pp.1411-1415.
Abstract - A crosstalk characterization model is proposed that shows the relationship between colors with and without crosstalk by using CIE tristimulus values. The proposed model is based on the 3D color measurement results of two different stereoscopic 3D televisions with shutter glasses, namely, a PDP...
Park, B.G., Lee, M.Y. and Yang, S.J., Samsung Electronics Co Ltd, 2012. Image display apparatus and method for correction chroma wrinkle. U.S. Patent 8,218,896.
Kim, S., Sim, J.Y. and Yang, S., 2012. Vision-based cleaning area control for cleaning robots. IEEE Transactions on Consumer Electronics, 58(2), pp.685-690.
Abstract - This paper provides a vision based HCI method for a user to command a cleaning robot to move to a specific location in home environment. Six hand poses are detected from a video sequence taken from a camera on the cleaning robot. AdaBoost based...
Kim, S.H., Yang, S.J., Park, R.H. and Lim, B.R., Samsung Electronics Co Ltd and Sogang University Industry University Cooperation Foundation, 2012. Image processing method and apparatus for contrast enhancement using intensity mapping. U.S. Patent 8,131,109.
Baek, Y.S., Kwak, Y. and Yang, S., 2012. Perceived Glossiness of Bumpy Surface.
Abstract - The perceived glossiness of 12 flat samples and 18 bumpy samples with various colors and gloss levels is estimated by 13 observers using magnitude estimation technique. Each sample is measured with the gloss-meter as well. It is found that bumpy surface shows lower measured...
Generalized minimal residual based parallel image restoration
National Research Foundation of Korea, NRF 2012~2015
Kim, S.H., Yang, S.J., Park, R.H., Lee, J.W., Lee, H.S., Kim, J.Y. and Lim, B.R., Samsung Electronics Co Ltd and Sogang University Industry University Cooperation Foundation, 2011. Method and apparatus for improving quality of composite video signal and method and apparatus for removing artifact of composite video signal. U.S. Patent 7,986,854.
Yang, S., 2011. Particle filtering based estimation of consistent motion and disparity with reduced search points. IEEE transactions on circuits and systems for video technology, 22(1), pp.91-104.
Abstract - A particle filtering based block-wise estimation method for estimation of motion in a video sequence and joint estimation of disparity and motion in a stereo video sequence is proposed. Parameters of motion and disparity of a block in a sequence are defined as a...
Lee, Y.H., Yang, S.J. and Hong, K.H., Samsung Electronics Co Ltd, 2011. Field sequential display apparatus that reduces color breakup and method thereof. U.S. Patent 7,952,549.
Yang, S., 2011. Reduced reference MPEG-2 picture quality measure based on ratio of DCT coefficients. Electronics Letters, 47(6), pp.382-383.
Abstract - A reduced reference picture quality measure is proposed for MPEG-2 compressed videos. The measure can be calculated from the ratio of three N tap one-dimensional discrete cosine transform (DCT) coefficients. The proposed measures are highly correlated to subjective test scores and can provide accurate...
Yang, S.J., Samsung Electronics Co Ltd, 2011. Motion adaptive image processing apparatus and method thereof. U.S. Patent 7,885,475.
Dynamic view generation for augmented reality
National Research Foundation of Korea, NRF 2009~2012
Han, S.H. and Yang, S.J., Samsung Electronics Co Ltd, 2010. Image conversion apparatus to perform motion compensation and method thereof. U.S. Patent 7,683,971.
Lee, Y.H. and Yang, S.J., Samsung Electronics Co Ltd, 2010. Image processing device capable of selecting field and method thereof. U.S. Patent 7,675,572.
A wavelet packet-based noise reduction algorithm of NTSC images using CVBS characteristics
Lim, B.R., Lee, H.S., Park, R.H. and Yang, S., 2009. A wavelet packet-based noise reduction algorithm of NTSC images using CVBS characteristics. IEEE Transactions on Consumer Electronics, 55(4), pp.2407-2415.
Lee, Y.H. and Yang, S.J., Samsung Electronics Co Ltd, 2009. Image conversion device and method. U.S. Patent 7,593,059.
Lee, Y.H. and Yang, S.J., Samsung Electronics Co Ltd, 2009. Image processing apparatus using judder-map and method thereof. U.S. Patent 7,499,102.
Yang, S.J. and Kwon, Y.J., Samsung Electronics Co Ltd, 2008. Apparatus to suppress artifacts of an image signal and method thereof. U.S. Patent 7,443,448.
Kim, D.H., Han, S.H., Yang, S.J. and Lee, A., Samsung Electronics Co Ltd, 2008. Method and apparatus for motion vector estimation using motion vectors of neighboring image blocks. U.S. Patent Application 11/830,026.
Han, S.H., Kim, D.H., Yang, S.J. and Lee, Y.H., Samsung Electronics Co Ltd, 2008. Apparatus for and method of estimating motion vector. U.S. Patent Application 11/742,632.
Yang, S.J., Samsung Electronics Co Ltd, 2008. De-interlacing apparatus with a noise reduction/removal device. U.S. Patent 7,324,160.
Park, Y.J., Oh, J.H., Kang, H. and Yang, S.J., Samsung Electronics Co Ltd, 2008. Contrast compensation apparatus and method thereof. U.S. Patent 7,359,573.
Hong, S.H., Park, R.H., Yang, S. and Kim, J.Y., 2008. Image interpolation using interpolative classified vector quantization. Image and Vision Computing, 26(2), pp.228-239.
Abstract - According to advances in digital imaging technology, interest in high-resolution (HR) images has been increased. Various methods that convert low-resolution (LR) images to HR ones have been presented. In this paper, to reduce the computational load we propose a vector quantization (VQ) based algorithm...
Kim, S.H., Park, R.H. and Yang, S., 2008, January. Superresolution of a region of interest using feature-based affine motion estimation. In 2008 Digest of Technical Papers-International Conference on Consumer Electronics (pp. 1-2). IEEE.
Abstract - This paper presents a low-computational superresolution method of a region of interest (ROI) using multiple low-resolution input images of the same scene. We extract the regions from the multiple input images that are similar to the selected ROI in the reference image and use...