• 한국인터넷방송통신학회논문지
• /
• 제23권6호
• /
• pp.89-95
• /
• 2023

인간 행동 인식 (Human action recognition, HAR) 기술은 스포츠 분석, 인간과 로봇 간의 상호작용, 대형 사이니지 콘텐츠 등의 애플리케이션에 활용되는 핵심 기술 중 하나이다. 본 논문에서는 몰입형 대형 사이니지 콘텐츠를 위한 STAGCN (Spatial temporal attention graph convolutional network) 기반 인간 행동 인식 시스템을 제안한다. STAGCN은 attention mechanism을 통해 스켈레톤 시퀀스의 시공간적 특징에 서로 다른 가중치를 부과하여, 동작 인식에 중요한 관절 및 시점을 고려할 수 있다. NTU RGB+D 데이터셋을 사용한 실험 결과, 제안된 시스템은 기존 딥러닝 모델들에 비해 높은 분류 정확도를 달성한 것을 확인했다.

• 로봇학회논문지
• /
• 제18권4호
• /
• pp.436-443
• /
• 2023

With the advancement of robot intelligence, the concept of man and unmanned teaming (MUM-T) has garnered considerable attention in military research. In this paper, we present a transformer-based architecture for predicting the health status of agents, with the help of multi-head attention mechanism to effectively capture the dynamic interaction between friendly and enemy forces. To this end, we first introduce a framework for generating a dataset of battlefield situations. These situations are simulated on a virtual simulator, allowing for a wide range of scenarios without any restrictions on the number of agents, their missions, or their actions. Then, we define the crucial elements for identifying the battlefield, with a specific emphasis on agents' status. The battlefield data is fed into the transformer architecture, with classification headers on top of the transformer encoding layers to categorize health status of agent. We conduct ablation tests to assess the significance of various factors in determining agents' health status in battlefield scenarios. We conduct 3-Fold corss validation and the experimental results demonstrate that our model achieves a prediction accuracy of over 98%. In addition, the performance of our model are compared with that of other models such as convolutional neural network (CNN) and multi layer perceptron (MLP), and the results establish the superiority of our model.

• Corrosion Science and Technology
• /
• 제22권5호
• /
• pp.377-386
• /
• 2023

Aircraft washing is performed periodically for corrosion control. Currently, the aircraft washing interval is qualitatively set according to the geographical conditions of each base. We developed a washing interval determination algorithm based on atmospheric corrosion environment monitoring data at the Republic of Korea Air Force (ROKAF) bases and United States Air Force (USAF) bases to determine the optimal interval. The main factors of the washing interval decision algorithm were identified through hierarchical clustering, sensitivity analysis, and analysis of variance, and criteria were derived. To improve the classification accuracy, we developed a washing interval decision model based on an artificial neural network (ANN). The ANN model was calibrated and validated using the atmospheric corrosion environment monitoring data and washing intervals of the USAF bases. The new algorithm returned a three-level washing interval, depending on the corrosion rate of steel and the results of the ANN model. A new base-specific aircraft washing interval was proposed by inputting the atmospheric corrosion environment monitoring results of the ROKAF bases into the algorithm.

• Journal of Information Processing Systems
• /
• 제19권6호
• /
• pp.730-744
• /
• 2023

The topic of this paper is the recognition of human activities using egocentric vision, particularly captured by body-worn cameras, which could be helpful for video surveillance, automatic search and video indexing. This being the case, it could also be helpful in assistance to elderly and frail persons for revolutionizing and improving their lives. The process throws up the task of human activities recognition remaining problematic, because of the important variations, where it is realized through the use of an external device, similar to a robot, as a personal assistant. The inferred information is used both online to assist the person, and offline to support the personal assistant. With our proposed method being robust against the various factors of variability problem in action executions, the major purpose of this paper is to perform an efficient and simple recognition method from egocentric camera data only using convolutional neural network and deep learning. In terms of accuracy improvement, simulation results outperform the current state of the art by a significant margin of 61% when using egocentric camera data only, more than 44% when using egocentric camera and several stationary cameras data and more than 12% when using both inertial measurement unit (IMU) and egocentric camera data.

• International Journal of Advanced Culture Technology
• /
• 제12권2호
• /
• pp.167-179
• /
• 2024

This study addresses the challenge of objectively evaluating the performance of early-stage startups amidst limited information and uncertainty. Focusing on companies selected by TIPS, a leading private sector-driven startup support policy in Korea, the research develops a new indicator to assess technological efficiency. By analyzing various input and output variables collected from Crunchbase and KIND (Korea Investor's Network for Disclosure System) databases, including technology use metrics, patents, and Crunchbase rankings, the study derives technological efficiency for TIPS-selected startups. A prediction model is then developed utilizing machine learning techniques such as Random Forest and boosting (XGBoost) to classify startups into efficiency percentiles (10th, 30th, and 50th). The results indicate that prediction accuracy improves with higher percentiles based on the technical efficiency index, providing valuable insights for evaluating and predicting startup performance in early markets characterized by information scarcity and uncertainty. Future research directions should focus on assessing growth potential and sustainability using the developed classification and prediction models, aiding investors in making data-driven investment decisions and contributing to the development of the early startup ecosystem.

• 스마트미디어저널
• /
• 제13권6호
• /
• pp.35-43
• /
• 2024

During the post-pandemic period, the interest in foreign language learning surged, leading to increased usage of language-learning apps. With the rising demand for these apps, analyzing app reviews becomes essential, as they provide valuable insights into user experiences and suggestions for improvement. This research focuses on extracting insights into users' opinions, sentiments, and overall satisfaction from reviews of HelloTalk, one of the most renowned language-learning apps. We employed topic modeling and emotion analysis approaches to analyze reviews collected from the Google Play Store. Several experiments were conducted to evaluate the performance of sentiment classification models with different settings. In addition, we identified dominant emotions and topics within the app reviews using feature importance analysis. The experimental results show that the Random Forest model with topics and emotions outperforms other approaches in accuracy, recall, and F1 score. The findings reveal that topics emphasizing language learning and community interactions, as well as the use of language learning tools and the learning experience, are prominent. Moreover, the emotions of 'admiration' and 'annoyance' emerge as significant factors across all models. This research highlights that incorporating emotion scores into the model and utilizing a broader range of emotion labels enhances model performance.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2024년도 춘계학술발표대회
• /
• pp.774-776
• /
• 2024

The effective implementation of advanced speech recognition (ASR) systems necessitates the deployment of sophisticated keyword spotting models that are both responsive and resource-efficient. The initial local detection of user interactions is crucial as it allows for the selective transmission of audio data to cloud services, thereby reducing operational costs and mitigating privacy risks associated with continuous data streaming. In this paper, we address these needs and propose utilizing the Whisper-Tiny model with fine-tuning process to specifically recognize keywords from google speech dataset which includes 65000 audio clips of keyword commands. By adapting the model's encoder and appending a lightweight classification head, we ensure that it operates within the limited resource constraints of local devices. The proposed model achieves the notable test accuracy of 92.94%. This architecture demonstrates the efficiency as on-device model with stringent resources leading to enhanced accessibility in everyday speech recognition applications.

• Journal of Korea Artificial Intelligence Association
• /
• 제1권1호
• /
• pp.11-16
• /
• 2023

Accurate hospital case modeling and prediction are crucial for efficient healthcare. In this study, we demonstrate the implementation of regression analysis methods in machine learning systems utilizing mathematical statics and machine learning techniques. The developed machine learning model includes Bayesian linear, artificial neural network, decision tree, decision forest, and linear regression analysis models. Through the application of these algorithms, corresponding regression models were constructed and analyzed. The results suggest the potential of leveraging machine learning systems for medical research. The experiment aimed to create an Azure Machine Learning Studio tool for the speedy evaluation of multiple regression models. The tool faciliates the comparision of 5 types of regression models in a unified experiment and presents assessment results with performance metrics. Evaluation of regression machine learning models highlighted the advantages of boosted decision tree regression, and decision forest regression in hospital case prediction. These findings could lay the groundwork for the deliberate development of new directions in medical data processing and decision making. Furthermore, potential avenues for future research may include exploring methods such as clustering, classification, and anomaly detection in healthcare systems.

• 정보관리학회지
• /
• 제17권4호
• /
• pp.27-45
• /
• 2000

인구센서스와 같은 국가 통계정보는 국가의 미래 투자계획과 정책수립을 위한 중요한 기초데이터이다. 그러나 데이터의 코딩과정이 모두 수작업으로 이루어지기 때문에 결과의 일관성 결여와 시간과 인력이 너무 많이 소요된다는 것 등이 문제점으로 지적되고 있다. 따라서 본 연구에서는 한국 산업표준 분류표에 근거한 자동코딩시스템을 개발하여 코딩과정을 수작업으로 처리할 때 발생하는 문제점을 해결하였다. 시스템의 지식베이스로는 학습이론을 사용하여 저자가 새로이 개발한 복수의 전거어 사전들을 활용하였다. 실험한 결과, 생성률은 99.5%를, 정확률은 83.3%라는 결과를 얻었다. 따라서 이 시스템은 실제 통계데이터의 자동코딩과정에 사용될 수 있으며, 국가 통계정보의 효율적 분석에 매우 유용하게 사용될 수 있을 것이다.

• 정보처리학회 논문지
• /
• 제13권5호
• /
• pp.217-220
• /
• 2024

지진 발생은 정확히 예측하기 어렵고, 이러한 무작위성을 갖는 사건에 대비하여 모든 건물에 내진 설계를 도입하는 것은 현실적으로 어려운 과제이다. 건물의 특징 분석을 통한 건물 손상 예측을 기반으로 건물의 취약점을 보완한다면, 내진 설계를 도입하지 않은 건물에서도 피해를 최소화할 수 있으므로 건물 손상 예측 모델의 효율성을 분석하는 연구가 필요하다. 본 논문에서는 2015년 네팔 대지진으로 인해 손상된 건물 데이터를 활용하여 Random Forest, Extreme Gradient Boosting, LightGBM, CatBoost 기계학습 분류 알고리즘을 사용하여 지진 피해 예측 모델의 정확도를 비교하였다.