• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.5
• /
• pp.682-687
• /
• 2022

In this paper, the research results on monolithic three-dimensional integrated-circuit (M3DICs) stacked with tunneling field effect transistors (TFETs) are introduced. Unlike metal-oxide-semiconductor field-effect transistors (MOSFETs), TFETs are designed differently from the layout of symmetrical MOSFETs because the source and drain of TFET are asymmetrical. Various monolithic 3D inverter (M3D-INV) structures and layouts are possible due to the asymmetric structure, and among them, a simple inverter structure with the minimum metal layer is proposed. Using the proposed M3D-INV, this M3D logic gates such as NAND and NOR gates by sequentially stacking TFETs are proposed, respectively. The simulation results of voltage transfer characteristics of the proposed M3D logic gates are investigated using mixed-mode simulator of technology computer aided design (TCAD), and the operation of each logic circuit is verified. The cell area for each M3D logic gate is reduced by about 50% compared to one for the two-dimensional planar logic gates.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.9
• /
• pp.1330-1339
• /
• 2022

3D object detection generally aims to detect relatively large data such as automobiles, buses, persons, furniture, etc, so it is vulnerable to small object detection. In addition, in an environment with limited resources such as embedded devices, it is difficult to apply the model because of the huge amount of computation. In this paper, the accuracy of small object detection was improved by focusing on local features using only one layer, and the inference speed was improved through the proposed knowledge distillation method from large pre-trained network to small network and adaptive quantization method according to the parameter size. The proposed model was evaluated using SUN RGB-D Val and self-made apple tree data set. Finally, it achieved the accuracy performance of 62.04% at mAP@0.25 and 47.1% at mAP@0.5, and the inference speed was 120.5 scenes per sec, showing a fast real-time processing speed.

• Journal of Sensor Science and Technology
• /
• v.30 no.4
• /
• pp.267-272
• /
• 2021

In this study, Ag was deposited to investigate its applicability as a surface-enhanced Raman scattering substrate after forming a grass-type black silicon structure through maskless reactive ion etching. Grass-structured black silicon with heights of 2 - 7 ㎛ was formed at radio-frequency (RF) power of 150 - 170 W. The process pressure was 250 mTorr, the O₂/SF₆ gas ratio was 15/37.5, and the processing time was 10 - 20 min. When the processing time was increased by more than 20 min, the self-masking of Si_xO_yF_z did not occur, and the black silicon structure was therefore not formed. Raman response characteristics were measured based on the Ag thickness deposited on a black silicon substrate. As the Ag thickness increased, the characteristic peak intensity increased. When the Ag thickness deposited on the black silicon substrate increased from 40 to 80 nm, the Raman response intensity at a Raman wavelength of 1507 / cm increased from 8.2 × 10³ to 25 × 10³ cps. When the Ag thickness was 150 nm, the increase declined to 30 × 10³ cps and showed a saturation tendency. When the RF power increased from 150 to 170 W, the response intensity at a 1507/cm Raman wavelength slightly increased from 30 × 10³ to 33 × 10³ cps. However, when the RF power was 200 W, the Raman response intensity decreased significantly to 6.2 × 10³ cps.

• Journal of Digital Policy
• /
• v.1 no.2
• /
• pp.61-69
• /
• 2022

The paper is a study to develop and verify a blood glucose level prediction model based on biosignals obtained from photoplethysmography (PPG) sensors, ICT technology and data. Blood glucose prediction used the MLP architecture of machine learning. The input layer of the machine learning model consists of 10 input nodes and 5 hidden layers: heart rate, heart rate variability, age, gender, VLF, LF, HF, SDNN, RMSSD, and PNN50. The results of the predictive model are MSE=0.0724, MAE=1.1022 and RMSE=1.0285, and the coefficient of determination (R²) is 0.9985. A blood glucose prediction model using bio-signal data collected from digital devices and machine learning was established and verified. If research to standardize and increase accuracy of machine learning datasets for various digital devices continues, it could be an alternative method for individual blood glucose management.

• Advanced Industrial SCIence
• /
• v.1 no.2
• /
• pp.24-29
• /
• 2022

Recently, smart farm research that applies IoT technology to various farms is being actively conducted to improve agricultural cooling power and minimize cost reduction. In particular, methods for automatically and remotely controlling environmental information data around smart farms through IoT devices are being studied. This paper proposes a processing model that can maintain an optimal growth environment by monitoring environmental information data collected from smart farms in real time based on machine learning. Since the proposed model uses machine learning technology, environmental information is grouped into multiple blockchains to enable continuous data collection through rich big data securing measures. In addition, the proposed model selectively (or binding) the collected environmental information data according to priority using weights and correlation indices. Finally, the proposed model allows us to extend the cost of processing environmental information to n-layer to a minimum so that we can process environmental information in real time.

• KIPS Transactions on Computer and Communication Systems
• /
• v.12 no.1
• /
• pp.1-8
• /
• 2023

It is impossible to overwrite on an already allocated page in SSDs, so whenever a write operation occurs a page replacement with a clean page is required. To resolve this problem, SSDs have an internal flash translation layer called FTL that maps logical pages managed by a file system of operating system to currently allocated physical pages. SSD pages discarded due to write operations must be recycled through initialization, but since the number of initialization times is limited the FTL provides a caching function to reduce the number of writes in addition to the page mapping function, which is a core function. In this study, we focus on the FTL cache methodologies reducing the number of page writes and analyze the related algorithms, and propose a write-only cache strategy. As a result of experimenting with the write-only cache using a simulator, it showed an improvement of up to 29%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.3976-3990
• /
• 2022

Currently, cross-attention is widely used in Siamese trackers to replace traditional correlation operations for feature fusion between template and search region. The former can establish a similar relationship between the target and the search region better than the latter for robust visual object tracking. But existing trackers using cross-attention only focus on rich semantic information of high-level features, while ignoring the appearance information contained in low-level features, which makes trackers vulnerable to interference from similar objects. In this paper, we propose a Multi-level Cross-attention Siamese network(MCSiam) to aggregate the semantic information and appearance information at the same time. Specifically, a multi-level cross-attention module is designed to fuse the multi-layer features extracted from the backbone, which integrate different levels of the template and search region features, so that the rich appearance information and semantic information can be used to carry out the tracking task simultaneously. In addition, before cross-attention, a target-aware module is introduced to enhance the target feature and alleviate interference, which makes the multi-level cross-attention module more efficient to fuse the information of the target and the search region. We test the MCSiam on four tracking benchmarks and the result show that the proposed tracker achieves comparable performance to the state-of-the-art trackers.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.11
• /
• pp.1477-1485
• /
• 2021

Recently, super-resolution has been intensively studied only on upscaling models with integer magnification. However, the need to expand arbitrary magnification is emerging in representative application fields of actual super-resolution, such as object recognition and display image quality improvement. In this paper, we propose a model that can support arbitrary magnification by using the weights of the existing integer magnification model. This model converts super-resolution results into the DCT spectral domain to expand the space for arbitrary magnification. To reduce the loss of high-frequency information in the image caused by the expansion by the DCT spectral domain, we propose a high-frequency attention network for arbitrary magnification so that this model can properly restore high-frequency spectral information. To recover high-frequency information properly, the proposed network utilizes channel attention layers. This layer can learn correlations between RGB channels, and it can deepen the model through residual structures.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.9
• /
• pp.399-406
• /
• 2023

Cloud computing has been evolved to support edge computing architecture that combines fog management layer with edge servers. The main reason why it is received much attention is low communication latency for real-time IoT applications. At the same time, various cloud task scheduling techniques based on artificial intelligence have been proposed. Artificial intelligence-based cloud task scheduling techniques show better performance in comparison to existing methods, but it has relatively high scheduling time. In this paper, we propose a deep learning-based dynamic scheduling with multi-agents supporting scalability in edge computing environments. The proposed method shows low scheduling time than previous artificial intelligence-based scheduling techniques. To show the effectiveness of the proposed method, we compare the performance between previous and proposed methods in a scalable experimental environment. The results show that our method supports real-time IoT applications with low scheduling time, and shows better performance in terms of the number of completed cloud tasks in a scalable experimental environment.

• International conference on construction engineering and project management
• /
• 2009.05a
• /
• pp.928-934
• /
• 2009

Although the evolution and deployment of information technologies will undoubtedly play an important role in the current construction industry, many engineers are still unsure of the economic value of using these technologies. Especially for the planning of a construction project, a collaboration system to utilize the whole resources is a essential tool for the successful outcome. A detailed, authoritative, and readily accessible information model is needed to enable engineers to make cost-effective decisions among established and innovative plan alternatives. Most engineers rely on limited private experiences when they create solutions or design alternatives. Initial planning is crucial for the success of the construction project. Most construction projects are done through collaboration of engineers who have different specialized knowledge. Information technologies can dramatically enhance the performance of the collaboration. For the information delivery, we need a mediator between engineers. Object-based 3-D models are useful for the communication and decision assistance for the intelligent project design. In this paper, basic guidelines for the 3-D design according to different construction processes are suggested. Adequate interoperability of 3-D objects from any CAD system is essential for the collaboration. Basic architectures of geometry models and their information layer were established to enable interoperability for design checks, estimation and simulation. A typical international project for roadway was chosen for the pilot project. 3-D GIS model was created and bridge information models were created considering several requirements for planning and decision making of the project. From the pilot test, the integrated construction project planning using 3-D information models was discussed and several guidelines were suggested.