The Intelligent Computing and Imaging Lab (ICL) conducts research in advanced medical imaging, including CT and PET/SPECT reconstruction, motion correction, and quantitative imaging. It focuses on generative AI and diffusion models, such as text-to-image/video generation, personalization, and efficient adaptation. The lab also investigates robust computer vision and image restoration, including super-resolution, semantic segmentation, denoising, and real-world visual enhancements. Research outcomes are published in major medical imaging and signal processing journals and presented at machine learning and computer vision conferences. ICL is part of the Department of Electrical and Computer Engineering and affiliated with the Interdisciplinary Program in Artificial Intelligence at Seoul National University. The lab is led by Se Young Chun.

지능형 컴퓨팅 및 이미징 연구실(ICL)은 CT 및 PET/SPECT 재구성, 모션 보정, 정량적 이미징을 포함한 첨단 의료 영상 연구를 수행한다. 텍스트-이미지/비디오 생성, 개인화, 효율적인 적응과 같은 생성적 AI 및 확산 모델에 중점을 둔다. 또한 초해상도, 시맨틱 분할, 노이즈 제거 및 실제 시각적 향상과 같은 강력한 컴퓨터 비전 및 이미지 복원을 연구한다. 연구 결과는 주요 의료 영상 및 신호 처리 저널에 게재되고 머신 러닝 및 컴퓨터 비전 컨퍼런스에서 발표된다. ICL은 서울대학교 전기컴퓨터공학부에 소속되어 있으며 인공지능 협동과정과 관련되어 있다. Se Young Chun이 연구실을 이끈다.

Fidelity-Preserving Zero-shot Diffusion Models for Highly Ill-posed Inverse Problems in Lensless Imaging

Robust Watermarks for Audio Diffusion Models by Quadrature Amplitude Modulation

Convergent Complex Quasi-Newton Proximal Methods for Gradient-Driven Denoisers in Compressed Sensing MRI Reconstruction

Clinical dementia rating classification using integrated vision and language information

A Simple Extension of Bayer Demosaicing For Light-Weight Non-Bayer Demosaicing

Sparse-view CBCT reconstruction using meta-learned neural attenuation field and hash-encoding regularization

Continual Test-time Adaptation for Robust Remote Photoplethysmography Estimation

Class Concept Representation from Contextual Texts for Training-Free Multi-Label Recognition

Image Quality Assessments By Leveraging Diverse Visual Tasks

Deep Network Regularization for Phase-based Magnetic Resonance Electrical Properties Tomography with Stein’s Unbiased Risk Estimator

On Epistemic Uncertainty of Visual Tokens for Object Hallucinations in Large Vision-Language Models

Continual Multiple Instance Learning with Enhanced Localization for Histopathological Whole Slide Image Analysis

PersonaCraft: Personalized Full-Body Image Synthesis for Multiple Identities from Single References Using 3D-Model-Conditioned Diffusion

Robust 3D-Masked Part-level Editing in 3D Gaussian Splatting with Regularized Score Distillation Sampling

GeoMan: Temporally Consistent Human Geometry Estimation using Image-to-Video Diffusion

MOST: MR reconstruction Optimization for multiple downStream Tasks via continual learning

Skrr: Skip and Re-use Text Encoder Layers for Memory Efficient Text-to-Image Generation

Localized Concept Erasure for Text-to-Image Diffusion Models Using Training-Free Gated Low-Rank Adaptation

Efficient Personalization of Quantized Diffusion Model without Backpropagation

Class Distribution-induced Attention Map for Open-vocabulary Semantic Segmentations

Concept Pinpoint Eraser for Text-to-image Diffusion Models via Residual Attention Gate

PersonaCraft: Personalized Full-body Image Synthesis for Multiple Identities from Single References Using 3D-Model-Conditioned Diffusion

Harmonization for a black-box deep learning model

Blind harmonization reinvented via 3D diffusion model

Few-Shot Arbitrary Super-Resolution Considering Consistency and Realness

Focusing on Representation of Multi-Head Attention for Open-Vocabulary Semantic Segmentation

Zero-Shot Diffusion Models for Demosaicing in Bayer and Non-Bayer Image Sensors

Object-Interactive Human Motion Generation via Affordance Reasoning

Robust Panoramic Localization via Hybrid Optimization

Training-Free Dynamics Score Control in Text-to-Video Diffusion Models via Wavelet-Based Latent Refinement

Self-Cascade Coordinate-guided Residual Diffusion Models for Arbitrary-Scale Perceptual Image Super-Resolution

Detailed texture reconstruction for multiple interacting humans using an iterative method

SpecArtDiff: ControlNet-Guided Fusion of Panoramic Visuals and Auditory Spectrograms

Closed-form Instance Unlearning for Explicit Distributions

Mitigating Hallucinations in Large Vision-Language Models via Norm-Guided Visual Filtering

Zero-shot Diffusion Model based Highly Ill-posed Inverse Problem Solver via Tweedie Variable Optimization with Total Variation

LaMoGen: Laban Movement-Guided Diffusion for Text-to-Motion Generation

Principled Reframing in Motion Transfer with Video Diffusion Models

Upsample What Matters: Region-Adaptive Latent Sampling for Accelerated Diffusion Transformers

Self-Cascaded Diffusion Models for Arbitrary-Scale Image Super-Resolution

On Geometrical Properties of Text Token Embeddings for Strong Semantic Binding in Text-to-Image Generation

Audio Data Management Device and Method Using A Generative AI Model

Method For Affordance Grounding and Apparatus Therefor

Device and Method for Extracting High-Dimensional Gene Fearures Based on Data Dimension Reduction

Method and Apparatus for Generating High-resolution Human-centric Scene Using Pretrained Diffusion Model

Deep Internal Learning: Deep Learning from a Single Input

Gradient-free Decoder Inversion in Latent Diffusion Models

INTRA: Interaction Relationship-aware Weakly Supervised Affordance Grounding

Contribution-based Low-Rank Adaptation with Pre-training Model for Real Image Restoration

BeyondScene: Higher-Resolution Human-Centric Scene Generation With Pretrained Diffusion

Adaptive Selection of Sampling-Reconstruction in Fourier Compressed Sensing

UL-VIO: Ultra-Lightweight Visual-Inertial Odometry with Noise Robust Test-Time Adaptation

On Exact Inversion of DPM-Solvers

Doubly Perturbed Task Free Continual Learning

AIM 2024 Challenge on UHD Blind Photo Quality Assessment

Ultra-lightweight Visual-Inertial Odometry with Noise Robust Test-time Adaptation

Uncertainty-Adaptive Selection of Sampling-Reconstruction for Fourier Compressed Sensing

Physics-Guided Deep Scatter Estimation by Weak Supervision for Quantitative SPECT

Selective Concept Erasing for Safe Diffusion Models

Zero-shot diffusion model using DDPG for lensless image restoration

Zero-Shot Super-Restoration Using Fourier-Guided Parallel Diffusion Model

Personalization of Abstract Concept in Text-to-Image Diffusion Model

Towards Personalization of Generative Models Via Zeroth-Order Optimization

Enhanced Depth Map Upsampling Using ControlNet

Efficient Representation of Heterogenous Features in Tabular Data: Unifying Structures for a Foundation Model

ACE-LCM: Attention-based Controllability and direct classifier-free guidance based quality Enhancing in few-step Latent Consistency Model

Improving Prompt Fidelity in Style-Aligned Image Generation via Attention Manipulation

Self-supervised denoising in PET considering real noise characteristics

Expert classifier ensemble based post-processing correction for unbiased scene graph generation

Short-term Object Interaction Anticipation with Disentangled Object Detection @ Ego4D Short Term Object Interaction Anticipation Challenge

Learning Method for Image Restoration Model and Apparatus for Performing The Same

Domain Adaptation Using Pose-Preserved Text-to-image Diffusion For 3D Generative Model

Towards accelerating model parallelism in distributed deep learning systems

PAIP 2020: Microsatellite Instability Prediction in Colorectal Cancer

Efficient Unified Demosaicing for Bayer and Non-Bayer Patterned Image Sensors

On the Robustness of Normalizing Flows for Inverse Problems in Imaging

Online Continual Learning on Hierarchical Label Expansion

PODIA-3D: Domain Adaptation of 3D Generative Model Across Large Domain Gap Using Pose-Preserved Text-to-Image Diffusion

DATID-3D: Diversity-Preserved Domain Adaptation Using Text-to-Image Diffusion for 3D Generative Model

All-in-one Image Restoration for Unknown Degradations Using Adaptive Discriminative Filters for Specific Degradations

Neural Diffeomorphic Non-Uniform B-Spline Flows

Fully Quantized Always-on Face Detector Considering Mobile Image Sensors

Domain adaptation from posteroanterior to anteroposterior X-ray radiograph classification via deep neural converter with label recycling

Deep neural network for cardiomegaly classification on AP-view X-ray

Detailed Human-Centric Text Description-Driven Large Scene Synthesis

DITTO-NeRF: Diffusion-based Iterative Text To Omni-directional 3D Model

All-in-one image quality improvement model providing method performing image quality restoration for multiple image quality degradation factors

Diversity-Preserved Domain Adaptation Using Text-to-Image Diffusion For 3D Generative Model

Ntire 2018 challenge on single image super-resolution: Methods and results

Multi-Temporal Recurrent Neural Networks For Progressive Non-Uniform Single Image Deblurring With Incremental Temporal Training

Ntire 2020 challenge on real-world image super-resolution: Methods and results

Task-aware variational adversarial active learning

MRI-based nonrigid motion correction in simultaneous PET/MRI

Training Deep Learning Based Denoisers without Ground Truth Data

PAIP 2019: Liver cancer segmentation challenge

Ntire 2019 challenge on real image super-resolution: Methods and results

Nonrigid PET motion compensation in the lower abdomen using simultaneous tagged‐MRI and PET imaging

Ntire 2019 challenge on real image denoising: Methods and results

A simple regularizer for B-spline nonrigid image registration that encourages local invertibility

Ntire 2019 image dehazing challenge report

Cardiac motion compensation and resolution modeling in simultaneous PET-MR: a cardiac lesion detection study

Improved quantitative 90 Y bremsstrahlung SPECT/CT reconstruction with Monte Carlo scatter modeling

Rethinking Deep Image Prior for Denoising

Extending Stein's unbiased risk estimator to train deep denoisers with correlated pairs of noisy images

Training deep learning based image denoisers from undersampled measurements without ground truth and without image prior

All-in-One Image Restoration for Unknown Degradations Using Adaptive Discriminative Filters for Specific Degradations

Correction for collimator-detector response in SPECT using point spread function template

A single multi-task deep neural network with post-processing for object detection with reasoning and robotic grasp detection

Se Young Chun

Professor

Intelligent Computing and Imaging Lab (ICL)

교수님 소개

천세영

연구실 소개

연구실 키워드

알럼나이가 간 회사

연구실 지원 방법

대우 조건

연도별 논문 인용수

저널 & 학회 논문 출판 수

연구실 논문 목록

Fidelity-Preserving Zero-shot Diffusion Models for Highly Ill-posed Inverse Problems in Lensless Imaging

Robust Watermarks for Audio Diffusion Models by Quadrature Amplitude Modulation

Convergent Complex Quasi-Newton Proximal Methods for Gradient-Driven Denoisers in Compressed Sensing MRI Reconstruction

연구실 보유 특허

대학별 입시끈

우리 연구실 모집

교수님 소개

천세영

연구실 키워드

알럼나이가 간 회사

연구실 지원 방법

대우 조건

연도별 논문 인용수

저널 & 학회 논문 출판 수

연구실 논문 목록

Fidelity-Preserving Zero-shot Diffusion Models for Highly Ill-posed Inverse Problems in Lensless Imaging

Robust Watermarks for Audio Diffusion Models by Quadrature Amplitude Modulation

Convergent Complex Quasi-Newton Proximal Methods for Gradient-Driven Denoisers in Compressed Sensing MRI Reconstruction

연구실 보유 특허

최근 3년간 연구실적 높은 연구실

대학별 입시끈

우리 연구실 모집