• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.8
• /
• pp.2764-2782
• /
• 2021

Existing city-level boundary nodes identification methods need to locate all IP addresses on the path to differentiate which IP is the boundary node. However, these methods are susceptible to time-delay, the accuracy of location information and other factors, and the resource consumption of locating all IPes is tremendous. To improve the recognition rate and reduce the locating cost, this paper proposes an algorithm for city-level boundary node identification based on bidirectional approaching. Different from the existing methods based on time-delay information and location results, the proposed algorithm uses topological analysis to construct a set of candidate boundary nodes and then identifies the boundary nodes. The proposed algorithm can identify the boundary of the target city network without high-precision location information and dramatically reduces resource consumption compared with the traditional algorithm. Meanwhile, it can label some errors in the existing IP address database. Based on 45,182,326 measurement results from Zhengzhou, Chengdu and Hangzhou in China and New York, Los Angeles and Dallas in the United States, the experimental results show that: The algorithm can accurately identify the city boundary nodes using only 20.33% location resources, and more than 80.29% of the boundary nodes can be mined with a precision of more than 70.73%.

• Molecules and Cells
• /
• v.45 no.7
• /
• pp.444-453
• /
• 2022

Multivalent macromolecular interactions underlie dynamic regulation of diverse biological processes in ever-changing cellular states. These interactions often involve binding of multiple proteins to a linear lattice including intrinsically disordered proteins and the chromosomal DNA with many repeating recognition motifs. Quantitative understanding of such multivalent interactions on a linear lattice is crucial for exploring their unique regulatory potentials in the cellular processes. In this review, the distinctive molecular features of the linear lattice system are first discussed with a particular focus on the overlapping nature of potential protein binding sites within a lattice. Then, we introduce two general quantitative frameworks, combinatorial and conditional probability models, dealing with the overlap problem and relating the binding parameters to the experimentally measurable properties of the linear lattice-protein interactions. To this end, we present two specific examples where the quantitative models have been applied and further extended to provide biological insights into specific cellular processes. In the first case, the conditional probability model was extended to highlight the significant impact of nonspecific binding of transcription factors to the chromosomal DNA on gene-specific transcriptional activities. The second case presents the recently developed combinatorial models to unravel the complex organization of target protein binding sites within an intrinsically disordered region (IDR) of a nucleoporin. In particular, these models have suggested a unique function of IDRs as a molecular switch coupling distinct cellular processes. The quantitative models reviewed here are envisioned to further advance for dissection and functional studies of more complex systems including phase-separated biomolecular condensates.

• Journal of the Semiconductor & Display Technology
• /
• v.21 no.4
• /
• pp.121-125
• /
• 2022

This paper deals with the development of a deep-learning-based robot that recognizes various types of stairs and performs a mission to go up to the target floor. The overall motion sequence of the robot is performed based on the ROS robot operating system, and it is possible to detect the shape of the stairs required to implement the motion sequence through rapid object recognition through YOLOv4 and Cuda acceleration calculations. Using the ROS operating system installed in Jetson Nano, a system was built to support communication between Arduino DUE and OpenCM 9.04 with heterogeneous hardware and to control the movement of the robot by aligning the received sensors and data. In addition, the web server for robot control was manufactured as ROS web server, and flow chart and basic ROS communication were designed to enable control through computer and smartphone through message passing.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.245-265
• /
• 2022

In response to problems such as insufficient extraction information, low detection accuracy, and frequent misdetection in the field of Thangka image defects, this paper proposes a YOLOv5 prediction algorithm fused with the attention mechanism. Firstly, the Backbone network is used for feature extraction, and the attention mechanism is fused to represent different features, so that the network can fully extract the texture and semantic features of the defect area. The extracted features are then weighted and fused, so as to reduce the loss of information. Next, the weighted fused features are transferred to the Neck network, the semantic features and texture features of different layers are fused by FPN, and the defect target is located more accurately by PAN. In the detection network, the CIOU loss function is used to replace the GIOU loss function to locate the image defect area quickly and accurately, generate the bounding box, and predict the defect category. The results show that compared with the original network, YOLOv5-SE and YOLOv5-CBAM achieve an improvement of 8.95% and 12.87% in detection accuracy respectively. The improved networks can identify the location and category of defects more accurately, and greatly improve the accuracy of defect detection of Thangka images.

• Journal of Broadcast Engineering
• /
• v.26 no.7
• /
• pp.868-876
• /
• 2021

Deploying deep neural network model training performs remarkable performance in the fields of Object detection and Instance segmentation. To train these models, features are first extracted from the input image using a backbone network. The extracted features can be reused by various tasks. Research has been actively conducted to serve various tasks by using these learned features. In this process, standardization discussions about encoding, decoding, and transmission methods are proceeding actively. In this scenario, it is necessary to analyze the response characteristics of features against various distortions that may occur in the data transmission or data compression process. In this paper, experiment was conducted to inject various distortions into the feature in the object recognition task. And analyze the mAP (mean Average Precision) metric between the predicted value output from the neural network and the target value as the intensity of various distortions was increased. Experiments have shown that features are more robust to distortion than images. And this points out that using the feature as transmission means can prevent the loss of information against the various distortions during data transmission and compression process.

• Smart Structures and Systems
• /
• v.30 no.2
• /
• pp.145-157
• /
• 2022

The design of safe and economical structures depends on the reliable live load from load survey. Live load surveys are traditionally conducted by randomly selecting rooms and weighing each item on-site, a method that has problems of low efficiency, high cost, and long cycle time. This paper proposes a deep learning-based method combined with Internet big data to perform live load surveys. The proposed survey method utilizes multi-source heterogeneous data, such as images, voice, and product identification, to obtain the live load without weighing each item through object detection, web crawler, and speech recognition. The indoor objects and face detection models are first developed based on fine-tuning the YOLOv3 algorithm to detect target objects and obtain the number of people in a room, respectively. Each detection model is evaluated using the independent testing set. Then web crawler frameworks with keyword and image retrieval are established to extract the weight information of detected objects from Internet big data. The live load in a room is derived by combining the weight and number of items and people. To verify the feasibility of the proposed survey method, a live load survey is carried out for a meeting room. The results show that, compared with the traditional method of sampling and weighing, the proposed method could perform efficient and convenient live load surveys and represents a new load research paradigm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.11
• /
• pp.3723-3737
• /
• 2022

Google Play is one of the largest Android phone app markets and it contains both free and paid apps. It provides a variety of categories for every target user who has different needs and purposes. The customer's rate every product based on their experience of apps and based on the average rating the position of an app in these arch varies. Fraudulent behaviors emerge in those apps which incorporate search rank maltreatment and malware proliferation. To distinguish the fraudulent behavior, a novel framework is structured that finds and uses follows left behind by fraudsters, to identify both malware and applications exposed to the search rank fraud method. This strategy correlates survey exercises and remarkably joins identified review relations with semantic and behavioral signals produced from Google Play application information, to distinguish dubious applications. The proposed model accomplishes 90% precision in grouping gathered informational indexes of malware, fakes, and authentic apps. It finds many fraudulent applications that right now avoid Google Bouncers recognition technology. It also helped the discovery of fake reviews using the reviewer relationship amount of reviews which are forced as positive reviews for each reviewed Google play the android app.

• Journal of the Korean Society of Industry Convergence
• /
• v.26 no.2_1
• /
• pp.217-223
• /
• 2023

With the rapid advancement of technologies, need for different research fields where this technology can be used is also increasing. One of the most researched topic in computer vision is object detection, which has widely been implemented in various fields which include healthcare, video surveillance and education. The main goal of object detection is to identify and categorize all the objects in a target environment. Specifically, methods of object detection consist of a variety of significant techniq ues, such as image processing and patterns recognition. Anomaly detection is a part of object detection, anomalies can be found various scenarios for example crowded places such as subway stations. An abnormal event can be assumed as a variation from the conventional scene. Since the abnormal event does not occur frequently, the distribution of normal and abnormal events is thoroughly imbalanced. In terms of public safety, abnormal events should be avoided and therefore immediate action need to be taken. When abnormal events occur in certain places, real time detection is required to prevent and protect the safety of the people. To solve the above problems, we propose a modified YOLOv5 object detection algorithm by implementing dilated convolutional layers which achieved 97% mAP50 compared to other five different models of YOLOv5. In addition to this, we also created a simple mobile application to avail the abnormal event detection on mobile phones.

• Korean Journal of Artificial Intelligence
• /
• v.11 no.3
• /
• pp.1-5
• /
• 2023

With recent advancements in artificial intelligence (AI) technology, more products based on AI are being launched and used. However, using AI safely requires an awareness of the potential risks it can pose. These concerns must be evaluated by experts and users must be informed of the results. In response to this need, many countries have implemented certification programs for products based on AI. In this study, we analyze several trends and differences in AI product certification programs across several countries and emphasize the importance of such programs in ensuring the safety and trustworthiness of products that include AI. To this end, we examine four international AI product certification programs and suggest methods for improving and promoting these programs. The certification programs target AI products produced for specific purposes such as autonomous intelligence systems and facial recognition technology, or extend a conventional software quality certification based on the ISO/IEC 25000 standard. The results of our analysis show that companies aim to strategically differentiate their products in the market by ensuring the quality and trustworthiness of AI technologies. Additionally, we propose methods to improve and promote the certification programs based on the results. These findings provide new knowledge and insights that contribute to the development of AI-based product certification programs.

• BMB Reports
• /
• v.56 no.2
• /
• pp.102-107
• /
• 2023

Genome editing using CRISPR-associated technology is widely used to modify the genomes rapidly and efficiently on specific DNA double-strand breaks (DSBs) induced by Cas9 endonuclease. However, despite swift advance in Cas9 engineering, structural basis of Cas9-recognition and cleavage complex remains unclear. Proper assembly of this complex correlates to effective Cas9 activity, leading to high efficacy of genome editing events. Here, we develop a CRISPR/Cas9-RAD51 plasmid constitutively expressing RAD51, which can bind to single-stranded DNA for DSB repair. We show that the efficiency of CRISPR-mediated genome editing can be significantly improved by expressing RAD51, responsible for DSB repair via homologous recombination (HR), in both gene knock-out and knock-in processes. In cells with CRISPR/Cas9-RAD51 plasmid, expression of the target genes (cohesin SMC3 and GAPDH) was reduced by more than 1.9-fold compared to the CRISPR/Cas9 plasmid for knock-out of genes. Furthermore, CRISPR/Cas9-RAD51 enhanced the knock-in efficiency of DsRed donor DNA. Thus, the CRISPR/Cas9-RAD51 system is useful for applications requiring precise and efficient genome edits not accessible to HR-deficient cell genome editing and for developing CRISPR/Cas9-mediated knockout technology.