Korea Advanced Institute of Science and Technology · 情報科学
Professor Jongmin Yu's research lab specializes in computer vision and deep learning with a focus on real-world applications in transportation safety and intelligent systems. The lab primarily investigates unsupervised and self-supervised representation learning for anomaly detection, including driver drowsiness detection, road surface defect recognition, and vehicle re-identification, with an emphasis on robustness under real-world variations such as lighting, weather, and data contamination. Key research directions include adversarial learning, spatio-temporal modeling, and domain generalization to enable reliable performance without reliance on large-scale annotated datasets. The lab develops innovative deep learning frameworks that enhance model generalization and detection accuracy in practical, open-world scenarios.
Figures are computed from collected data and may differ slightly.
We propose a condition-adaptive representation learning framework for driver drowsiness detection based on a 3D-deep convolutional neural network. The proposed framework consists of four models: spatio-temporal representation learning, scene condition understanding, feature fusion, and drowsiness detection. Spatio-temporal representation learning extracts features that can describe motions and appearances in video simultaneously. Scene condition understanding classifies the scene conditions rela
We present adversarial event prediction (AEP), a novel approach to detecting abnormal events through an event prediction setting. Given normal event samples, AEP derives the prediction model, which can discover the correlation between the present and future of events in the training step. In obtaining the prediction model, we propose adversarial learning for the past and future of events. The proposed adversarial learning enforces AEP to learn the representation for predicting future events and
Ensuring traffic safety is crucial, which necessitates the detection and prevention of road surface defects. As a result, there has been a growing interest in the literature on the subject, leading to the development of various road surface defect detection methods. The methods for detecting road defects can be categorised in various ways depending on the input data types or training methodologies. The predominant approach involves image-based methods, which analyse pixel intensities and surface
In the past few years, the performance of road defect detection has been remarkably improved thanks to advancements in various studies on computer vision and deep learning. Although large-scale and well-annotated datasets enhance the performance of detecting road defects to some extent, it is still challengeable to derive a model which can perform reliably for various road conditions in practice, because it is intractable to construct a dataset considering diverse road conditions and defect patt
The key challenge of unsupervised vehicle re-identification (Re-ID) is learning discriminative features from unlabelled vehicle images. Numerous methods using domain adaptation have achieved outstanding performance, but those methods still need a labelled dataset as a source domain. This paper addresses an unsupervised vehicle Re-ID method, which no need any types of a labelled dataset, through a Self-supervised Metric Learning (SSML) based on a feature dictionary. Our method initially extracts
In this paper, we propose Normality-Calibrated Autoencoder (NCAE), which can boost anomaly detection performance on the contaminated datasets without any prior information or explicit abnormal samples in the training phase. The NCAE adversarially generates high confident normal samples from a latent space having low entropy and leverages them to predict abnormal samples in a training dataset. NCAE is trained to minimise reconstruction errors in uncontaminated samples and maximise reconstruction
This paper presents a Gaussian-Poisson mixture model (GPMM) which can reflect a frequency of event occurrence, for detecting anomaly of crowd behaviours. GPMM exploits the complementary information of both a statistics of crowd behaviour patterns and a count of the observed behaviour, and we learn the statistics of normal crowd behaviours for behaviours that occur frequently in the past by placing different weights, depending on the frequency occur. GPMM implicitly accounts for the motion patter
Determination of model complexity is a challenging issue to solve computer vision problems using restricted boltzmann machines (RBMs). Many algorithms for feature learning depend on cross-validation or empirical methods to optimize the number of features. In this work, we propose an learning algorithm to find the optimal model complexity for the RBMs by incrementing the hidden layer. The proposed algorithm is composed of two processes: 1) determining incrementation necessity of neurons and 2) co
In this paper, we address face verification and person re-identification tasks under the unconstrained condition. Both tasks under the unconstrained condition are difficult since the testing dataset can contain the identities not appeared in training dataset. To overcome the difficulty, we propose deep discriminative representation learning (DDRL) to learn a discriminative representation which can cover not only trained representation but the appearance of images which are not trained. DDRL can
Recently, vehicle re-identification methods based on deep learning constitute remarkable achievement. However, this achievement requires large-scale and well-annotated datasets. In constructing the dataset, assigning globally available identities (Ids) to vehicles captured from a great number of cameras is labour-intensive, because it needs to consider their subtle appearance differences or viewpoint variations. In this paper, we propose camera-tracklet-aware contrastive learning (CTACL) using t
Open papers in the app to read, cite, and organize with AI.
