We are a passionate team of researchers at Korea University, Republic of Korea, dedicated to advancing the frontiers of signal learning. Our current focus lies in foundation model-based efficient deep learning methods, positioning us at the forefront of upcoming technological innovations. Join us as we explore and contribute to the latest breakthroughs in this dynamic and evolving field!

우리는 신호 학습의 최전선을 발전시키기 위해 헌신하는 대한민국 고려대학교의 열정적인 연구팀입니다. 우리의 현재 초점은 파운데이션 모델 기반의 효율적인 딥러닝 방법에 있으며, 이는 우리를 다가오는 기술 혁신의 최전선에 위치시킵니다. 이 역동적이고 진화하는 분야의 최신 혁신을 탐구하고 기여하는 데 함께 하십시오!

Knowledge Tailoring: Bridging the Teacher-Student Gap in Semantic Segmentation

Ranked Entropy Minimization for Continual Test-Time Adaptation

Sensor Fusion-based Autoencoder Feature Distillation for 3D Object Detection

Bridging Domain Spaces for Unsupervised Domain Adaptation

Semantic Prompting with Image Token for Continual Learning

Channel and Spatial Enhancement Network for Human Parsing

Dual Prototype-driven Objectness Decoupling for Cross-Domain Object Detection in Urban Scene

SRIL: Selective Regularization for Class-Incremental Learning

Stay Focus on Object: Cross-Domain Detection using Domain-Invariant Object Representation

D3T: Distinctive Dual-Domain Teacher Zigzagging Across RGB-Thermal Gap for Domain-Adaptive Object Detection

Switching Temporary Teachers for Semi-supervised Semantic Segmentation

ORC: Network Group-based Knowledge Distillation using Online Role Change

Class Relationship-based Knowledge Distillation for Efficient Human Parsing

itKD: Interchange Transfer-based Knowledge Distillation for 3D Object Detection

Weighted Alignment-based Multi-Source Domain Adaptation for Object Detection

Contrastive Vicinal Space for Unsupervised Domain Adaptation

CDGNet: Class Distribution Guided Network for Human Parsing

Colored Point Cloud Registration by Depth Filtering

Densely Guided Knowledge Distillation using Multiple Teacher Assistants

Self-Supervised Visual Learning by Variable Playback Speeds Prediction of a Video

FixBi: Bridging Domain Spaces for Unsupervised Domain Adaptation

Low bit-based Convolutional Neural Network for One-class Object Detection

자율 이동체를 위한 연속 테스트 시점 적응 방법 및 그 장치

다중 파운데이션 모델 융합 기반 휴먼 파싱 방법 및 그 장치

지식 증류 기반 3차원 객체 탐지 모델의 경량화를 수행하는 전자장치 및 경량화 방법

Method and Device for Human Parsing

학생의 지식에 따라 교사의 지식을 커스터마이징 하는 지식 증류 방법 및 장치

이미지 토큰을 이용한 시맨틱 프롬프트 기반의 인공지능 모델 학습 방법 및 시스템

공간 분포를 이용한 휴먼 파싱 방법 및 장치

Method and Apparatus for Unsupervised Domain Adaptation

인공지능을 이용한 객체 검출을 수행하기 위한 방법 및 장치

영상의 회전 특성 학습 방법 및 디바이스

카테고리에 따른 도메인 판별자를 통한 멀티 모달 도메인 간 객체 검출 방법

도메인 적응형 객체 검출을 위해 RGB 열 간격에 대한 지그재그 학습을 제공하는 변별적 이중 도메인 교사

인공지능 알고리즘 모듈을 학습시키기 위한 방법 및 장치

인공지능을 이용한 시맨틱 세그멘테이션을 수행하기 위한 방법 및 장치

도메인 불변 지식을 통한 비지도 도메인 적응 기반의 객체 감지 모델 학습

선택적 정규화를 통한 클래스 증분 학습

인공지능을 이용한 객체 검출 방법 및 장치

비지도 도메인 적응 방법 및 장치

Generative Adversarial Network (GAN)-based Sytstem for Generating Color Image From Edge Image

경량화 네트워크를 이용한 휴먼 파싱 방법 및 시스템

Electronic Apparatus for Weight Reduction of 3D Object Detection Model based on Knowledge Distillation and Weight Reduction Method Therefor

지식 증류를 활용한 그룹 기반의 학습을 수행하는 전자장치 및 방법

Method and Apparatus for Measuring Interaction Force Based On Sequential Images using Attention Network

구형 물체를 이용한 다시점 카메라 위치 보정 장치 및 방법

딥러닝 학습을 위한 데이터 처리 방법 및 그 장치

이종 도메인 데이터 간의 변환을 수행하는 GAN의 학습 방법 및 장치

에지(edge) 이미지으로부터 컬러(color) 이미지를 생성하기 위한 GAN(generative adversarial networks) 기반 시스템

지식 증류를 이용한 학생 네트워크 학습용 시스템 및 학습 방법

순차적인 이미지 기반의 상호작용력 측정 방법 및 장치

전자 장치 및 그 제어 방법

영상 기반 힘 예측 시스템 및 그 방법

객체 검출 알고리즘의 학습 방법 및 장치

Method and apparatus of recognizing face using 2nd-order independent component analysis (ICA)/principal component analysis (PCA)

Face Recognition Method and Apparatus using Component-based Face Descriptor

Method and Apparatus of Recognizing Face using Component-based 2nd-order Principal Component Analysis (PCA)/Independent Component Analysis (ICA)

Image Retrieval Method and Apparatus using Iterative Matching

Method and Apparatus for Extracting Feature Vector used for Face Recognition and Retrieval

Face Image Verification Method and Apparatus using A Kernel Based Discriminant Analysis with Local Binary Pattern (LBP)

Apparatus and Method for Generating Shape Model of Object and Apparatus and Method for Automatically Searching for Feature Points of Object Employing the Same

Face Recognition Apparatus and Method using PCA Learning per Subgroup

Method and Apparatus to Extracting Facial Features from Image Containing Face

Method and System for Image Restoration using Frequency-based Image Model

Image Processing Apparatus of Robot System and Method and Computer-readable Medium Thereof

3D Object Tracking Method and Apparatus

Method and Apparatus for Estimating Face Position in 3 Dimensions

Face Recognition Apparatus and Method

Face Recognition Apparatus and Method for Controlling the Same

Method and Apparatus to Generate Object Descriptor Using Extended Curvature Gabor Filter

IMAGE PROCESSING APPARATUS AND METHOD THEREOF

SURGICAL ROBOT SYSTEM AND METHOD OF CONTROLLING THE SAME

LIVENESS TESTING METHODS AND APPARATUSES AND IMAGE PROCESSING METHODS AND APPARATUSES

X-RAY IMAGING APPARATUS AND METHOD OF CONTROLLING THE SAME

Method and apparatus for detecting biometric region for user authentication

METHODS OF TRACING REGIONS OF INTEREST, RADIOGRAPHIC APPARATUSES, METHODS OF CONTROLLING THE RADIOGRAPHIC APPARATUSES, AND RADIOGRAPHY METHODS

Apparatus and Method for Retrieving Face Images using Combined Component Descriptors

Apparatus and Method for Processing Image based on Layers

Scalable Face Recognition Method and Apparatus based on Complementary Features of Face Image

Method, Medium, and System Recognizing A Face, and Method, Medium, and System Extracting Features from A Facial Image

Method and Apparatus for Processing Video Data

Method and Apparatus for Face Recognition using Gender Information

Method and Apparatus for Face Recognition Using Extended Gabor Wavelet Features

Method and Apparatus for Generating Face Descriptor Using Extended Local Binary Patterns, and Method and Apparatus for Face Recognition using Extended Local Binary Patterns

Method and Apparatus for Calculating Similarity of Face Image, Method and Apparatus for Retrieving face Image, and Method of Synthesizing Face Image

Method, Medium, and Apparatus with Object Descriptor Generation using Curvature Gabor Filter

Robot System and Method and Computer-readable Medium Controlling The Same

METHOD AND APPARATUS FOR POSE RECOGNITION

IMAGE PROCESSING APPARATUS AND IMAGE PROCESSING METHOD

DISPLAY APPARATUSES AND CONTROL METHODS THEREOF

ENDOSCOPE AND IMAGE PROCESSING APPARATUS USING THE SAME

SURGICAL TROCARS AND IMAGE ACQUISITION METHOD USING THE SAME

AUGMENTED REALITY IMAGE DISPLAY SYSTEM AND SURGICAL ROBOT SYSTEM COMPRISING THE SAME

ENDOSCOPE SYSTEM AND CONTROL METHOD THEREOF

METHOD AND APPARATUS OF IDENTIFYING USER USING FACE RECOGNITION

Method and Apparatus for Facial Recognition

Multi-modal Fusion Method For User Authentication and User Authentication Method

Augmented reality image display system and surgical robot system comprising the same

Fixbi: Bridging domain spaces for unsupervised domain adaptation

Surgical robot system and method of controlling the same

Densely guided knowledge distillation using multiple teacher assistants

Method and apparatus of recognizing face using component-based 2nd-order principal component analysis (PCA)/independent component analysis (ICA)

Component-based LDA face description for image retrieval and MPEG-7 standardisation

Face recognition system using multiple face model of hybrid Fourier feature under uncontrolled illumination variation

Face recognition method and apparatus using component-based face descriptor

Scalable face recognition method and apparatus based on complementary features of face image

Switching temporary teachers for semi-supervised semantic segmentation

Independent component analysis in a local facial residue space for face recognition

Cdgnet: Class distribution guided network for human parsing

Multiple face model of hybrid fourier feature for large face image set

Inferring interaction force from visual information without using physical force sensors

SVD face: illumination-invariant face representation

2D barcode detection using images for drone-assisted inventory management

Method and apparatus for adjusting face pose

Self-supervised spatio-temporal representation learning using variable playback speed prediction

Apparatus and method for retrieving face images using combined component descriptors

itkd: Interchange transfer-based knowledge distillation for 3d object detection

Jaemin Na

Tae-Kyun (T-K) Kim

Hyunwoo Kim

Hyeon Cho

Hyung Jin Chang

Jae-Joon Han

Junmo Kim

Soo Chul Lim

Wonjun Kim

Heechul Jung

Josef Kittler

Dongyoon Han

Ouk Choi

Haitao Wang

Biao Wang

ByungIn Yoo (유병인)

Hanseok Ko

Dinh Phat Do

Jisu Han

Seon-Min Rhee

Wonjun Hwang

An augmented reality image display system may be imple mented together with a surgical robot system. The surgical robot system may include a slave system performing a surgical operation, a master system controlling the surgical operation of the slave system, an imaging system generating a virtual image of the inside of a patient's body, and an augmented reality image display system including a camera capturing a real image having a plurality of markers attached to the patient's body or a human body model. The aug mented reality image system may include an augmented reality image generator which detects the plurality of mark ers in the real image, estimates the position and gaze direction of the camera using the detected markers, and generates an augmented reality image by overlaying a region of the virtual image over the real image, and a display which displays the augmented reality image.

Unsupervised domain adaptation (UDA) methods for learning domain invariant representations have achieved remarkable progress. However, most of the studies were based on direct adaptation from the source domain to the target domain and have suffered from large domain discrepancies. In this paper, we propose a UDA method that effectively handles such large domain discrepancies. We introduce a fixed ratio-based mixup to augment multiple intermediate domains between the source and target domain. From the augmented-domains, we train the source-dominant model and the target-dominant model that have complementary characteristics. Using our confidence-based learning methodologies, eg, bidirectional matching with high-confidence predictions and self-penalization using low-confidence predictions, the models can learn from each other or from its own results. Through our proposed methods, the models gradually transfer domain knowledge from the source to the target domain. Extensive experiments demonstrate the superiority of our proposed method on three public benchmarks: Office-31, Office-Home, and VisDA-2017.

(57) ABSTRACT A Surgical robot system includes a slave system to perform a Surgical operation on a patient and an imaging system that includes an image capture unit including a plurality of cameras to acquire a plurality of affected area images, an image generator detecting an occluded region in each of the affected area images acquired by the plurality of cameras, removing the occluded region therefrom, warping each of the affected area images from which the occluded region is removed, and matching the affected area images to generate a final image, and a controller driving each of the plurality of cameras of the image capture unit to acquire the plurality of affected area images and inputting the acquired plurality of affected area images to the image generator to generate a final image.

With the success of deep neural networks, knowledge distillation which guides the learning of a small student network from a large teacher network is being actively studied for model compression and transfer learning. However, few studies have been performed to resolve the poor learning issue of the student network when the student and teacher model sizes significantly differ. In this paper, we propose a densely guided knowledge distillation using multiple teacher assistants that gradually decreases the model size to efficiently bridge the large gap between the teacher and student networks. To stimulate more efficient learning of the student network, we guide each teacher assistant to every other smaller teacher assistants iteratively. Specifically, when teaching a smaller teacher assistant at the next step, the existing larger teacher assistants from the previous step are used as well as the teacher network. Moreover, we design stochastic teaching where, for each mini-batch, a teacher or teacher assistants are randomly dropped. This acts as a regularizer to improve the efficiency of teaching of the student network. Thus, the student can always learn salient distilled knowledge from the multiple sources. We verified the effectiveness of the proposed method for a classification task using CIFAR-10, CIFAR-100, and ImageNet. We also achieved significant performance improvements with various backbone architectures such as ResNet, WideResNet, and VGG.

(57) ABSTRACT A method and apparatus for recognizing and searching for a face using 2nd-order independent component analysis (ICA) are provided. The method for describing feature points uses 2nd-order ICA d to describe a facial image space and improve recognition performance in various illumination conditions. According to the method and apparatus, use of pose or illumination invariant face descriptor enables retrieval of human faces and authentication of a specific individual.

We propose a method of face description for facial image retrieval from a large data base and for MPEG-7 (Moving Picture Experts Group) standardisation. The novel descriptor is obtained by decomposing a face image into several components and then combining the component features. The decomposition combined with LDA (Linear Discriminant Analysis) provides discriminative facial descriptions that are less sensitive to light and pose changes. Each facial component is represented in its Fisher space and another LDA is then applied to compactly combine the features of the components. To enhance retrieval accuracy further, a simple pose classification and transformation technique is performed, followed by recursive matching. Our algorithm has been developed to deal with the problem of face image retrieval from huge databases such as those found in Internet environments. Such retrieval requires a …

The authors present a robust face recognition system for large-scale data sets taken under uncontrolled illumination variations. The proposed face recognition system consists of a novel illumination-insensitive preprocessing method, a hybrid Fourier-based facial feature extraction, and a score fusion scheme. First, in the preprocessing stage, a face image is transformed into an illumination-insensitive image, called an “integral normalized gradient image,” by normalizing and integrating the smoothed gradients of a facial image. Then, for feature extraction of complementary classifiers, multiple face models based upon hybrid Fourier features are applied. The hybrid Fourier features are extracted from different Fourier domains in different frequency bandwidths, and then each feature is individually classified by linear discriminant analysis. In addition, multiple face models are generated by plural normalized face images …

Provided are a face recognition apparatus and method in which a facial image is divided into facial component images. The apparatus includes a component division unit that divides an input facial image into a plurality of facial component images; a face descriptor generating unit that generates face descriptors using a transform matrix corre sponding to the respective facial component images, the face descriptors being characteristic vectors; a registered face descriptor database (DB) that stores registered face descrip tors; and an authentication unit that authenticates the input facial image by comparing face descriptors for the input facial image input from the face descriptor generating unit with the registered face descriptors and providing predeter mined weights corresponding to each facial component to the comparison results of each facial component.

[0005] With the development of the information society, the importance of identi? cation technology to identify indi viduals has rapidly groWn, and more research has been conducted on biometric technology for protecting computer based personal information and identifying individuals using the characteristics of the human body. In particular, face recognition, Which is a type of biometric technique, uses a non-contact method to identify individuals, and is thus deemed more convenient and more competitive than other biometric techniques such as? ngerprint recognition and iris recognition Which require users to behave in a certain Way to be recognized. Face recognition is a core technique for multimedia database searching, and is Widely used in various application? elds such as moving picture summariZation using face information, identity certi? cation, human computer interface (HCI) image searching, and secu …

The teacher-student framework, prevalent in semi-supervised semantic segmentation, mainly employs the exponential moving average (EMA) to update a single teacher's weights based on the student's. However, EMA updates raise a problem in that the weights of the teacher and student are getting coupled, causing a potential performance bottleneck. Furthermore, this problem may become more severe when training with more complicated labels such as segmentation masks but with few annotated data. This paper introduces Dual Teacher, a simple yet effective approach that employs dual temporary teachers aiming to alleviate the coupling problem for the student. The temporary teachers work in shifts and are progressively improved, so consistently prevent the teacher and student from becoming excessively close. Specifically, the temporary teachers periodically take turns generating pseudo-labels to train a student model and maintain the distinct characteristics of the student model for each epoch. Consequently, Dual Teacher achieves competitive performance on the PASCAL VOC, Cityscapes, and ADE20K benchmarks with remarkably shorter training times than state-of-the-art methods. Moreover, we demonstrate that our approach is model-agnostic and compatible with both CNN-and Transformer-based models. Code is available at https://github. com/naver-ai/dual-teacher.

In this paper, we propose an Independent Component Analysis (ICA) based face recognition algorithm, which is robust to illumination and pose variation. Generally, it is well known that the first few eigenfaces represent illumination variation rather than identity. Most Principal Component Analysis (PCA) based methods have overcome illumination variation by discarding the projection to a few leading eigenfaces. The space spanned after removing a few leading eigenfaces is called the “residual face space”. We found that ICA in the residual face space provides more efficient encoding in terms of redundancy reduction and robustness to pose variation as well as illumination variation, owing to its ability to represent non-Gaussian statistics. Moreover, a face image is separated into several facial components, local spaces, and each local space is represented by the ICA bases (independent components) of its …

The objective of human parsing is to partition a human in an image into constituent parts. This task involves labeling each pixel of the human image according to the classes. Since the human body comprises hierarchically structured parts, each body part of an image can have its sole position distribution characteristic. Probably, a human head is less likely to be under the feet, and arms are more likely to be near the torso. Inspired by this observation, we make instance class distributions by accumulating the original human parsing label in the horizontal and vertical directions, which can be utilized as supervision signals. Using these horizontal and vertical class distribution labels, the network is guided to exploit the intrinsic position distribution of each class. We combine two guided features to form a spatial guidance map, which is then superimposed onto the baseline network by multiplication and concatenation to distinguish the human parts precisely. We conducted extensive experiments to demonstrate the effectiveness and superiority of our method on three well-known benchmarks: LIP, ATR, and CIHP databases.

The face recognition system based on the only single classifier considering the restricted information can not guarantee the generality and superiority of performances in a real situation. To challenge such problems, we propose the hybrid Fourier features extracted from different frequency bands and multiple face models. The hybrid Fourier feature comprises three different Fourier domains; merged real and imaginary components, Fourier spectrum and phase angle. When deriving Fourier features from three Fourier domains, we define three different frequency bandwidths, so that additional complementary features can be obtained. After this, they are individually classified by Linear Discriminant Analysis. This approach makes possible analyzing a face image from the various viewpoints to recognize identities. Moreover, we propose multiple face models based on different eye positions with a same image size, and it …

In this paper, we present an interaction force estimation method that uses visual information rather than that of a force sensor. Specifically, we propose a novel deep learning-based method utilizing only sequential images for estimating the interaction force against a target object, where the shape of the object is changed by an external force. The force applied to the target can be estimated by means of the visual shape changes. However, the shape differences in the images are not very clear. To address this problem, we formulate a recurrent neural network-based deep model with fully-connected layers, which models complex temporal dynamics from the visual representations. Extensive evaluations show that the proposed learning models successfully estimate the interaction forces using only the corresponding sequential images, in particular in the case of three objects made of different materials, a sponge, a PET bottle, a human arm, and a tube. The forces predicted by the proposed method are very similar to those measured by force sensors.

In this letter, we propose a novel method to extract illumination-invariant features for face recognition and verification under varying illuminations. Inspired by the fact that normalized coefficients of the singular value decomposition (SVD) are insensitive to different illumination conditions, we exploit a simple, yet powerful scheme for describing underlying structures of faces, so-called SVD face. In contrast to previous approaches still suffering from the loss of details, our SVD face greatly preserves textures of the original image based on the relaxation of SVD coefficients. Theoretical analysis shows that our SVD face is an illumination-invariant measure and has an ability to discover meaningful components (e.g., eyes, mouth, etc.) of face images while suppressing the effect of various illuminations. Experimental results on both Yale B and our illuminated face (IF) datasets demonstrate that the SVD face is effective for face …

Drone-assisted inventory management is attractive for companies with large warehouses and factories. Additionally, there is an interest in a novel method that automatically detects target barcodes using a IR-based camera, which enables efficient drone path planning and results in reducing power consumption. In this paper, we propose an efficient detection framework which determines the localizations of 2D barcodes. Many regional proposals of 2D barcodes are reduced to a few candidate regions according to the distance information between the drone and the target 2D barcode. Visual features of the selected candidate regions are extracted by LBP and HOG methods, respectively. To gain discriminant power for classification, SVM is used at the end of the procedure. The final detection region is determined by a weighted sum-based score fusion method. To validate the performance of the proposed method, we …

ABSTRACT A method and an apparatus for adjusting a pose in a face image are provided. The method of adjusting a pose in a face image involves detecting two-dimensional (2D) landmarks from a 2D face image, positioning three-dimensional (3D) landmarks in a 3D face model by determining an initial pose of the 3D face model based on the 2D landmarks, updating the 3D landmarks by iteratively adjusting a pose and a shape of the 3D face model, and adjusting a pose in the 2D face image based on the updated 3D landmarks.

We propose a self-supervised learning method by predicting the variable playback speeds of a video. Without semantic labels, we learn the spatiotemporal representation of the video by leveraging the variations in the visual appearance according to different playback speeds under the assumption of temporal coherence. To learn the spatio-temporal variations in the entire video, we have not only predicted a single playback speed but also generated clips of various playback speeds with randomized starting points. We then train a 3D convolutional network by solving the formulation that sorts the shuffled clips by their playback speed. In this case, the playback speed includes both forward and reverse directions; hence the visual representation can be successfully learned from the directional dynamics of the video. We also propose a novel layer-dependable temporal group normalization method that can be applied to 3D convolutional networks to improve the representation learning performance where we divide the temporal features into several groups and normalize each one using the different corresponding parameters. We validate the effectiveness of the proposed method by fine-tuning it to the action recognition task. The experimental results show that the proposed method outperforms state-of-the-art self-supervised learning methods in action recognition.

Disclosed herein is an apparatus and method for retrieving face images using combined component descriptors. The apparatus of the present invention includes an image divi Sion unit for dividing an input image into facial components, a first Linear Discriminant Analysis transformation unit for LDA transforming the divided facial components into com ponent descriptors of the facial components, a vector Syn thesis unit for Synthesizing the transformed component descriptors into a Single vector, a Second Generalized Dis criminant Analysis transformation unit for GDA transform ing the Single vector into a Single face descriptor, and a Similarity determination unit. The Similarity determination unit determines similarities between an input query face image and face images Stored in a face image database by comparing a face descriptor of the input query face image with face descriptors of the face images Stored in the …

Point-cloud based 3D object detectors recently have achieved remarkable progress. However, most studies are limited to the development of network architectures for improving only their accuracy without consideration of the computational efficiency. In this paper, we first propose an autoencoder-style framework comprising channel-wise compression and decompression via interchange transfer-based knowledge distillation. To learn the map-view feature of a teacher network, the features from teacher and student networks are independently passed through the shared autoencoder; here, we use a compressed representation loss that binds the channel-wised compression knowledge from both student and teacher networks as a kind of regularization. The decompressed features are transferred in opposite directions to reduce the gap in the interchange reconstructions. Lastly, we present an head attention loss to match the 3D object detection information drawn by the multi-head self-attention mechanism. Through extensive experiments, we verify that our method can train the lightweight model that is well-aligned with the 3D point cloud detection task and we demonstrate its superiority using the well-known public datasets; eg, Waymo and nuScenes.

Full Professor

신호학습연구실

교수님 소개

황원준

연구실 소개

연구실 키워드

알럼나이가 간 회사

연구실 지원 방법

대우 조건

연도별 논문 인용수

저널 & 학회 논문 출판 수

연구실 논문 목록

Knowledge Tailoring: Bridging the Teacher-Student Gap in Semantic Segmentation

Ranked Entropy Minimization for Continual Test-Time Adaptation

Sensor Fusion-based Autoencoder Feature Distillation for 3D Object Detection

연구실 보유 특허

대학별 입시끈

우리 연구실 모집

교수님 소개

황원준

연구실 키워드

알럼나이가 간 회사

연구실 지원 방법

대우 조건

연도별 논문 인용수

저널 & 학회 논문 출판 수

연구실 논문 목록

Knowledge Tailoring: Bridging the Teacher-Student Gap in Semantic Segmentation

Ranked Entropy Minimization for Continual Test-Time Adaptation

Sensor Fusion-based Autoencoder Feature Distillation for 3D Object Detection

연구실 보유 특허

최근 3년간 연구실적 높은 연구실

대학별 입시끈

우리 연구실 모집