Tae-Sun Chung

• Ajou University

Vision Foundation Model (VFM) such as the Segment Anything Model (SAM) and Contrastive Language-Image Pre-training Model (CLIP) has shown promising performance for segmentation and detection tasks. However, although SAM excels in fine-grained segmentation, it faces major challenges when applying it to semantic-aware segmentation. While CLIP exhibit...

Accurate detection and diagnosis of brain tumors at early stages is significant for effective treatment. While numerous methods have been developed for tumor detection and classification, several rely on traditional techniques, often resulting in suboptimal performance. In contrast, AI-based deep learning techniques have shown promising results, co...

In recent years, significant advancements have been made in spatiotemporal sequence prediction, with PredRNN++ emerging as a powerful model due to its superior ability to capture complex temporal dependencies. However, the current unidirectional nature of PredRNN++ limits its ability to fully exploit the temporal information inherent in many real-w...

In recent years, significant advancements have been made in spatiotemporal sequence prediction, with PredRNN++ emerging as a powerful model due to its superior ability to capture complex temporal dependencies. However, the current unidirectional nature of PredRNN++ limits its ability to fully exploit the temporal information inherent in many real-w...

Conventional deduplication systems face critical challenges such as excessive write amplification, high read/write latency, and sub-optimal storage utilization. These limitations often undermine the performance benefits of deduplication by slowing down I/O acknowledgements due to amplified deduplication I/Os, excessive data chunk replication, and s...

Conventional deduplication systems face critical challenges such as excessive write amplification, high read/write latency, and sub-optimal storage utilization. These limitations often undermine the performance benefits of deduplication by slowing down I/O acknowledgements, due to amplified deduplication I/Os, excessive data chunk replication and s...

The rise in IoT (Internet of Things) devices poses a significant security challenge. Maintaining privacy and ensuring anonymity within the system is a sought-after feature with inevitable trade-offs, such as scalability and increased complexity, making it incredibly challenging to handle. To tackle this, we introduce our proposed work for managing...

The age of pervasive computing has initiated a boom in the development of adaptive context-aware learning environments (ACALEs), i.e., systems that are capable of detecting a learner’s context and providing adaptive learning services based on this context. Many of the existing educational systems were developed as standalone applications for specif...

The rise in IoT (Internet of Things) devices poses a significant security challenge. Maintaining privacy and ensuring anonymity within the system is a sought-after feature with inevitable trade-offs, such as scalability and increased complexity, making it incredibly challenging to handle. To tackle this, we introduce our proposed work for managing...

With the increase in IoT (Internet of Things) devices comes an inherent challenge of security. In the world today, privacy is the prime concern of every individual. Preserving one’s privacy and keeping anonymity throughout the system is a desired functionality that does not come without inevitable trade-offs like scalability and increased complexit...

Sign Language Production (SLP) refers to the task of translating textural forms of spoken language into corresponding sign language expressions. Sign languages convey meaning by means of multiple asynchronous articulators, including manual and non‐manual information channels. Recent deep learning‐based SLP models directly generate the full‐articula...

Machine translation requires that source and target sentences have identical semantics. Previous neural machine translation (NMT) models have implicitly achieved this requirement using cross-entropy loss. In this paper, we propose a sentence Semantic-aware Machine Translation model (SaMT) which explicitly addresses the issue of semantic similarity...

Unsupervised contrastive learning of sentence embedding has been a recent focus of researchers. However, issues such as unreasonable division of positive and negative samples and poor data enhancement leading to text semantic changes still exist. We propose an optimized data augmentation method that combines contrastive learning’s data augmentation...

Various components are involved in the end-to-end path of data transfer. Protecting data integrity from failures in these intermediate components is a key feature of big data transfer tools. Although most of these components provide some degree of data integrity, they are either too expensive or inefficient in recovering corrupted data. This proble...

Infectious diseases are always alarming for the survival of human life and are a key concern in the public health domain. Therefore, early diagnosis of these infectious diseases is a high demand for modern-era healthcare systems. Novel general infectious diseases such as coronavirus are infectious diseases that cause millions of human deaths across...

The proliferation of current and next-generation mobile and sensing devices has increased at an alarming rate. With these state-of-the-art devices, the global positioning system ( GPS ) has made remote sensing and location tracking more viable. One such query is the All Nearest Neighbor ( ANN ) query, which extracts and returns all data objects t...

Computational science simulations produce huge volumes of data for scientific research organizations. Often, this data is shared by data centers distributed geographically for storage and analysis. Data corruption in the end-to-end route of data transmission is one of the major challenges in distributing the data geographically. End-to-end integrit...

With the current advancement in the Internet, there has been a growing demand for building intelligent and smart systems that can efficiently address the detection of health-related problems on social media, such as the detection of depression and anxiety. These types of systems, which are mainly dependent on machine learning techniques, must be ab...

The usage of a smartphone while driving has been declared a global portent and has been admitted as a leading cause of crashes and accidents. Numerous solutions, such as Android Auto and CarPlay, are used to facilitate for the drivers by minimizing driver distractions. However, these solutions restrict smartphone usage, which is impractical in real...

Large scale machine learning (ML) and deep learning (DL) platforms face challenges when integrated with deduplication enabled storage clusters. In the quest to achieve smart and efficient storage utilization, removal of duplicate data introduces bottlenecks since deduplication alters the I/O transaction layout of the storage system. Therefore, it i...

The increasing trend of GPS-enabled smartphones has led to the tremendous usage of Location-Based Service applications. In the past few years, a significant amount of studies have been conducted to process All nearest neighbor (ANN) queries. An ANN query on a road network extracts and returns all the closest data objects for all query objects. Most...

Parkinson’s disease (PD) is one of the most common and serious neurological diseases. Impairments in voice have been reported to be the early biomarkers of the disease. Hence, development of PD diagnostic tool will help early diagnosis of the disease. Additionally, intelligent system developed for binary classification of PD and healthy controls ca...

Face image analysis is one among several important cues in computer vision. Over the last five decades, methods for face analysis have received immense attention due to large scale applications in various face analysis tasks. Face parsing strongly benefits various human face image analysis tasks inducing face pose estimation. In this paper we propo...

Various scientific research organizations generate several petabytes of data per year through computational science simulations. These data are often shared by geographically distributed data centers for data analysis. One of the major challenges in distributed environments is failure; hardware, network, and software might fail at any instant. Thus...

The popular approach for several natural language processing tasks involves deep neural networks, and in particular, recurrent neural networks (RNNs) and convolutional neural networks (CNNs). While RNNs can capture the dependency in a sequence of arbitrary length, CNNs are suitable for extracting position-invariant features. In this study, a state-...

Acceleration change index (ACI) is a fast and easy to understand heart rate variability (HRV) analysis approach used for assessing cardiac autonomic control of the nervous systems. The cardiac autonomic control of the nervous system is an example of highly integrated systems operating at multiple time scales. Traditional single scale based ACI did...

The task of referring expression comprehension (REC) is to localise an image region of a specific object described by a natural language expression, and all existing REC methods assume that the object described by the referring expression must be located in the given image. However, this assumption is not correct in some real applications. For exam...

The growing collection of scientific data in various web repositories is referred to as Scientific Big Data, as it fulfills the four “V’s” of Big Data–-volume, variety, velocity, and veracity. This phenomenon has created new opportunities for startups; for instance, the extraction of pertinent research papers from enormous knowledge repositories us...

In this paper, a hybrid scheme of Dynamic wireless charging (DWC) for electric vehicles EV(s) is proposed to resolve this issue in a network topological infrastructure. The proposed hybrid scheme uses different parameters to allow DWC in EVs. The network infrastructure was established through an enhanced destination sequential distance vector (Enha...

Flash memory prevalence has reached greater extents with its performance and compactness capabilities. This enables it to be easily adopted as storage media in various portable devices which includes smart watches, cell-phones, drones, and in-vehicle infotainment systems to mention but a few. To support large flash storage in such portable devices,...

Surface electromyography (sEMG) signals acquired with linear electrode array are useful in analyzing muscle anatomy and physiology. Most algorithms for signal processing, detection, and estimation require adequate quality of the input signals, however, multi-channel sEMG signals are commonly contaminated due to several noise sources. The sEMG signa...

In the modern world, people face an explosion of information and difficulty to find the right choice of their interest. Nowadays, people show interest in online shopping to meet their demands increasingly. For researchers and students, finding and buying the desired books from online shops is very tedious work. Recently Recommender System is an exc...

In this paper, we study the problem of string similarity search to retrieve in a database all strings similar to a query string within a given threshold. To measure the similarity between strings, we use edit distance. Many algorithms have been proposed under a filtering-and-verification framework to solve the problem. To reduce the overhead of edi...

Human face image analysis is an active research area within computer vision. In this paper we propose a framework for face image analysis, addressing three challenging problems of race, age, and gender recognition through face parsing. We manually labeled face images for training an end-to-end face parsing model through Deep Convolutional Neural Ne...

Generative adversarial networks (GANs) have shown significant progress in generating highquality visual samples, however they are still well known both for being unstable to train and for the problem of mode collapse, particularly when trained on data collections containing a diverse set of visual objects. In this paper, we propose an Adaptive k-st...

Many prior works have investigated on how to increase the job processing performance and energy efficient computing in large scale clusters. However, they employ serialized scheduling approaches encompassed with task straggler “hunting” techniques which launches speculative tasks after detecting slow tasks. These slow tasks are detected through nod...

Although flash memory solid state drives (FSSDs) outperform traditional hard disk drives (HDDs), their performance still fails to cope up with the perennial doubling speeds of microprocessors, regardless of the available high bandwidth. To alleviate this bottleneck, many semiconductor companies, such as Intel, Micron, Samsung, and Hynix have alread...

The way in which jobs are scheduled is critical to achieve high job processing performance in large scale data clusters. Most existing scheduling mechanism employs a First-In First-Out, serialized approach encompassed with task straggler hunting techniques which launches speculative tasks after detecting slow tasks. This is often achieved through t...

Artificial intelligence (AI) based Consumer Electronic (CE) devices generate massive amounts of accumulated hot data. Persisting this data into such devices’ storage media like the Universal flash Storage (USF) becomes an issue for flash storage system to manage hot data optimally. Consequently, efforts have been made to support many CE application...

The internal nature of flash memory technology, makes its performance highly dependent on workload characteristics causing poor performance on random writes. To solve this, Demand-based Flash Translation Layer (DFTL) which selectively caches page-level address mappings, was proposed. DFTL exploits temporal locality in workloads and when low, high c...

A top- k spatial keyword (T k Sk) query ranks objects based on the distance to the query location and textual relevance to the query keywords. Several solutions have been proposed for top- k spatial keyword queries. However, most of the studies focus on Euclidean space or only investigate the snapshot queries where both the query and data object ar...

A lot of previous works on Map-Reduce improved job completion performance through implementing additional instrumentation modules which collects system level information for making scheduling decisions. However the extra instrumentation may not scale well with increasing number of task-trackers. To this end, we design PADS, a lightweight scheduler...

In terms of system reliability, data recovery is a crucial capability. The lack of data recovery leads to the permanent loss of valuable data. This paper aims at improving data recovery in flash-based storage devices where extremely poor data recovery is shown. For this, we focus on garbage collection that determines the life span of data which hav...

Hadoop MapReduce is one of the widely used frameworks for large scale data processing. However , the scheduling model for MapReduce was designed with homogeneous clusters in mind which assumes same compute capabilities for all task-tracker nodes. Such scheduling does not scale well given a heterogeneous cluster environment. The LATE scheduler was d...

Flash memory has a variety of advantages such as the better performance compare to hard disk, shock resistance, small size, and so on. Therefore the use ratio of flash memory is being increased. But if the power-off failure is occurred, flash memory storage systems may loss mapping information. So in this paper, we propose power-off recovery scheme...

Most vendors of e-commerce applications deploy the cache memory to deliver the web objects to clients faster. However, they face many problems in dealing with the cache memory due to limited resources and dynamic access patterns. As a result, we need to efficiently manage the cache memory by evicting the unused data. The performance of cache manage...

This article offers an empirical exploration on the efficient use of word-level convolutional neural networks (word-CNN) for large-scale text classification. Generally, the word-CNNs are difficult to train on large-scale datasets as the size of word embedding dramatically increases as the size of vocabulary increases. In order to handle this issue,...

Flash memory has been a dominant storage medium in consumer electronics. Its wide deployment results from the remarkable progress in garbage collection mechanism which resolves the inherent limitation, not-in-place-update property, of flash memory. In this paper, we propose a new garbage collection management scheme, called selective-delay garbage...

Even though flash memory Solid State Drives (FSSDs) outperformed traditional Hard Disk Drives (HDDs), they are still failing to reduce performance gap between microprocessors and storage in computer systems regardless of available high bandwidth. To alleviate this, we propose implementing PCM as main memory in SSDs to replace flash memory. In parti...

Given a query location and a set of query keywords, a top-k spatial keyword query rank objects based on the distance to the query location and textual relevance to the query keywords. Several solutions have been proposed for top-k spatial keyword queries in Euclidean space. However, few algorithms study top-k keyword queries in undirected road netw...

In recent years, flash memory has been widely used in embedded devices and enterprise computing environments because of its many advantages, which includes shock resistance, low energy consumption, non-volatile nature and high I/O speed. However, flash memory hardware also possesses characteristics such as erase-before write, a limited number of er...

Text mining is a mechanism to find information by extracting resources from natural language. Compared with structured data in databases, text is unstructured and difficult to be dealt with for analyzing. Additionally, it is tedious tasks for users to identify accurate data. Text mining algorithm is similar to data mining, except that it processes...

A reverse k nearest neighbor (RkNN) query retrieves all the data points that have q as one of their k closest points. In recent years, considerable research has been conducted into monitoring reverse k nearest neighbor queries. In this paper, we study the problem of continuous reverse nearest neighbor queries where both the query object q and data...

In road networks, k-range nearest neighbor (k-RNN) queries locate the k-closest neighbors for every point on the road segments, within a given query region defined by the user, based on the network distance. This is an important task because the user's location information may be inaccurate; furthermore, users may be unwilling to reveal their exact...

Top-k spatial preference queries rank objects based on the score of feature objects in their spatial neighborhood. Top-k preference queries are crucial for a wide range of location based services such as hotel browsing and apartment searching. In recent years, a lot of research has been conducted on processing of top-k spatial preference queries in...

Flash Translation Layer (FTL) performs virtual-to-physical address translations and hides the erase-before-write characteristics of Flash. Pure page mapped FTL, which maintains page-level address mappings, is known as the most efficient FTL. However, its huge SRAM requirement to load the entire mapping table limited adoption of its use. In order to...

In this paper, we focus on read operations in flash memory, which have received less attention than write operations. To reduce read latency, we propose a read-aware dynamic allocation mechanism for multichannel solid-state devices. The proposed mechanism enables read operations to be executed immediately by reserving the resources of channels, pac...

Recently, flash memory is widely used as a non-volatile storage for embedded applications such as smart phones, MP3 players, digital cameras and so on. The software layer called flash translation layer (FTL) becomes more important since it is a key factor in the overall flash memory system performance. Many researchers have proposed FTL algorithms...

To avoid traffic accidents, drivers must constantly be aware of nearby vehicles. Unfortunately, nearby vehicles often go unnoticed because of various obstacles such as other vehicles, buildings, or poor weather. In this paper, we study Moving range k-nearest neighbor (MRkNN) queries as a tool for continuously monitoring nearby moving objects. A sim...

Top-k spatial preference query ranks objects based on the score of feature objects in their spatial neighborhood. Top-k preference queries are crucial for wide range of location based services such as hotel browsing and apartment searching; several algorithms have been proposed to process them in Euclidean space. Although, few algorithms study top-...

Reliable erasing of data from storage devices is a critical component of secure data management and is well understood for magnetic disks. However, flash memory has unusual electronic limitations that make in-place updating impossible. Many secure deletion techniques have been proposed to improve both information security and erasing reliability. T...

Recently developed query processors for complex events filter data streams to match specific patterns and transform streaming data into meaningful information. To specify the patterns to be matched, complex event processing languages adopt a negation to present an event that should not appear at a specific position. A negation engenders more expres...

Loop optimization, made of a sequence of loop transformations, plays an important role in performance improvement in data centric applications. Programs using flash memory are no exception to this, but, under certain conditions careless applications of specific loop transformations might cause unexpected results, due to the characteristics of flash...

Location privacy is a major obstacle in the ubiquitous deployment of mobile and pervasive computing services. In this study, we present a new approach for preserving the trajectory privacy of moving \(k\)-nearest neighbor (M\(k\)NN) queries in road networks. Several location anonymization algorithms have been proposed for providing location privacy...

In this paper, we investigate a new approach to moving kk-nearest neighbor (MkkNN) queries in directed and dynamic road networks, where each road segment has a particular orientation and its travel time changes depending on traffic conditions. An MkkNN query continuously finds the kk nearest neighbors (NNs) of a moving query object. Most existing s...

In this paper, we investigate the problem of computing the safe exit points of moving range queries in directed road networks where each road segment has a particular orientation. The safe exit point of query object q indicates the point at which the safe region and non-safe region of q meet. A safe region indicates a region where the query result...

Flash memory is rapidly deployed as data storage for embedded and tablet PCs due to its shock resistance, fast access, and low power consumption. However, it has some intractable characteristics such as erase-before-write, asymmetric read/write/erase speed, and limited number of write/erase cycles. Due to these hardware limitations, the magnetic di...

This paper presents a novel B-tree storage scheme that reduces the number of time-consuming erase operations on flash memory-based systems. The scheme adopts group-based round robin fashion and in-block logging technique for efficiently maintaining the B- tree nodes in the heavy-update workloads. Experimental results show that the proposed scheme i...

Major digital consumer electronics such as smartphones and tablet PCs are equipped with flash memory because of its many advantages. However, its distinguishing characteristics, including erase-before-update, asymmetric read/write/erase cost and limited number of erase cycles, make it necessary to reconsider existing storage access designs in order...

Flash memory is used widely in the data storage market, particularly low-price MultiLevel Cell (MLC) flash memory, which has been adopted by large-scale storage systems despite its low performance. To overcome the poor performance of MLC flash memory, a system architecture has been designed to optimize chip-level parallelism. This design increases...

In this paper, we investigate how traditional loop transformations affect flash memory performance. To do so, we propose cost models of loop structures in BAST (Block Associative Sector Translation) on flash memory. The proposed cost models reflect not only distinctive characteristics of flash memory but also loop structures, making these new model...

The design of flash memory systems for smart devices differs significantly from traditional storage systems, because most updates involve the random data. A previously proposed algorithm known as Switchable Address Translation (SAT) enhances the performance of multimedia storage devices; however, it exhibits low space utilization and executes inten...

In this paper, we study the processing of top-k spatial preference queries in road networks. A top-k spatial preference query retrieves a ranked list of the k best data objects based on the scores (e.g., qualities) of feature objects in their spatial neighborhoods. Several solutions have been proposed for top-k spatial preference queries in Euclide...

Flash memory is rapidly being deployed as a data storage medium for embedded systems and tablet computers due to its shock resistance, fast access, and low power consumption, etc. However, it has some intractable characteristics, such as erase-before-write, asymmetric read/write/erase speed, and a limited number of write/erase cycles. Due to these...

The continuous improvement in hardware design and advances in wireless communication have enabled the deployment of various wireless applications. Wireless sensor network applications become essential tools for monitoring the activity and evolution of our surrounding environment. However, the wireless sensor nodes are highly resource constrained in...

Query processing in road networks has been studied extensively in recent years. However, the processing of moving queries in road networks has received little attention. In this paper, we introduce a new algorithm called the Safe Exit Algorithm SEA, which can efficiently compute the safe exit points of a moving nearest neighbor NN query on road net...

Owing to its desirable characteristics, flash memory has become attractive to different hardware vendors as a primary choice for data storage. However, because of a limited number of block-erase lifecycles, it has become mandatory to redesign the existing approaches to maximize the flash memory lifetime. Wear-leveling is a mechanism that helps to e...

The polymers of intrinsic microporosity have gained the attention as one of the potential materials for membrane gas separation. The contorted ladder-like structure in polymers of intrinsic microporosity, specifically PIM-1 possesses high permeability but with moderate selectivity for O2/N2, CO2/N2 and CO2/CH4 separation. We report here the most co...

Given two positive parameters k and r, a constrained k-nearest neighbor CkNN query returns the k closest objects within a network distance r of the query location in road networks. In terms of the scalability of monitoring these CkNN queries, existing solutions based on central processing at a server suffer from a sudden and sharp rise in server lo...

Wireless sensor network applications become essential tools for monitoring the activity and evolution of our surrounding environment. Since, the key characteristics of wireless sensor networks are to handle very sensitive real-time tasks like surveillance and target tracking, the performance of processing devices within sensor nodes highly impact t...

Flash memory is being rapidly deployed as data storage for embedded devices such as PDAs, MP3 players, mobile phones and digital cameras due to its low electronic power, non-volatile storage, high performance, physical stability and portability. The most prominent characteristic of flash memory is that prewritten data can only be dynamically update...

Solid State Drive (SSD) is rapidly deployed as data storage for embedded and tablet computers due to its shock resistance, fast access, and low power consumption. However, it has some intractable characteristics such as erase-before-write, asymmetric read/write/erase speed, and limited number of write/erase cycles. Due to these hardware characteris...

Flash memory becomes ideal storage media for small size consumer electronics and embedded systems as well as for large size multimedia applications because of its attractive features like fast access speed, shock resistance, high reliability and further more it's rapidly increasing capacity. However, flash needs expensive erase operation for rewrit...

Discovery of association rules among the large number of item sets is considered as an important aspect of data mining. The ever increasing demand of finding pattern from large data enhances the association rule mining. Researchers developed a lot of algorithms and techniques for determining association rules. The main problem is the generation of...

Recently, flash memory is widely used as a non-volatile storage for embedded applications such as cellular phones, mp3 players, digital cameras, and so on. The software layer called FTL (flash translation layer) becomes more important since it is a key factor in the overall flash memory system performance. Many researchers have proposed FTL algorit...

Flash memory becomes ideal storage media for small size embedded systems as well as for large size multimedia applications because of its attractive features like fast access speed, shock resistance, high reliability and further more it's rapidly increasing capacity. However, flash needs expensive erase operations for rewriting data on same memory...