• Smart Media Journal
• /
• v.12 no.1
• /
• pp.9-16
• /
• 2023

Recently, a convolutional neural network (CNN) based system is being developed to overcome the limitations of human resources in the apple quality classification of farmhouse. However, since convolutional neural networks receive only images of the same size, preprocessing such as sampling may be required, and in the case of oversampling, information loss of the original image such as image quality degradation and blurring occurs. In this paper, in order to minimize the above problem, to generate a image patch based graph of an original image and propose a random walk-based positional encoding method to apply the graph transformer model. The above method continuously learns the position embedding information of patches which don't have a positional information based on the random walk algorithm, and finds the optimal graph structure by aggregating useful node information through the self-attention technique of graph transformer model. Therefore, it is robust and shows good performance even in a new graph structure of random node order and an arbitrary graph structure according to the location of an object in an image. As a result, when experimented with 5 apple quality datasets, the learning accuracy was higher than other GNN models by a minimum of 1.3% to a maximum of 4.7%, and the number of parameters was 3.59M, which was about 15% less than the 23.52M of the ResNet18 model. Therefore, it shows fast reasoning speed according to the reduction of the amount of computation and proves the effect.

• Korean Journal of Artificial Intelligence
• /
• v.9 no.1
• /
• pp.1-7
• /
• 2021

In this paper, We focused the issue of creating a socially problematic nurse schedule. The nurse schedule should be prepared in consideration of three shifts, appropriate placement of experienced workers, the fairness of work assignment, and legal work standards. Because of the complex structure of the nurse schedule, which must reflect various requirements, in most hospitals, the nurse in charge writes it by hand with a lot of time and effort. This study attempted to automatically create an optimized nurse schedule based on legal labor standards and fairness. We developed an I/O Q-Learning algorithm-based model based on Python and Web Application for automatic nurse schedule. The model was trained to converge to 100 by creating an Fairness Indicator Score(FIS) that considers Labor Standards Act, Work equity, Work preference. Manual nurse schedules and this model are compared with FIS. This model showed a higher work equity index of 13.31 points, work preference index of 1.52 points, and FIS of 16.38 points. This study was able to automatically generate nurse schedule based on reinforcement Learning. In addition, as a result of creating the nurse schedule of E hospital using this model, it was possible to reduce the time required from 88 hours to 3 hours. If additional supplementation of FIS and reinforcement Learning techniques such as DQN, CNN, Monte Carlo Simulation and AlphaZero additionally utilize a more an optimized model can be developed.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.136-136
• /
• 2021

In this study, a deep convolutional neural network (DCNN) model is proposed for short-term precipitation forecasting using weather radar-based images. The DCNN model is a combination of convolutional neural networks, autoencoder neural networks, and U-net architecture. The weather radar-based image data used here are retrieved from competition for rainfall forecasting in Korea (AI Contest for Rainfall Prediction of Hydroelectric Dam Using Public Data), organized by Dacon under the sponsorship of the Korean Water Resources Association in October 2020. This data is collected from rainy events during the rainy season (April - October) from 2010 to 2017. These images have undergone a preprocessing step to convert from weather radar data to grayscale image data before they are exploited for the competition. Accordingly, each of these gray images covers a spatial dimension of 120×120 pixels and has a corresponding temporal resolution of 10 minutes. Here, each pixel corresponds to a grid of size 4km×4km. The DCNN model is designed in this study to provide 10-minute predictive images in advance. Then, precipitation information can be obtained from these forecast images through empirical conversion formulas. Model performance is assessed by comparing the Score index, which is defined based on the ratio of MAE (mean absolute error) to CSI (critical success index) values. The competition results have demonstrated the impressive performance of the DCNN model, where the Score value is 0.530 compared to the best value from the competition of 0.500, ranking 16^th out of 463 participating teams. This study's findings exhibit the potential of applying the DCNN model to short-term rainfall prediction using weather radar-based images. As a result, this model can be applied to other areas with different spatiotemporal resolutions.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.11a
• /
• pp.252-255
• /
• 2019

본 논문에서 Equirectangular projection(ERP) 영상을 행동 인식 모델에 입력하기전 제안하는 전처리를 통하여 성능을 향상시키는 것을 보인다. ERP 영상의 특성상 행동 인식을 하는데 불필요한 영역이 일반적인 2D 카메라로 촬영한 영상보다 많다. 또한 행동 인식은 사람이 Object of Interest(OOI)이다. 따라서 객체 인식모델로 인간 객체를 인식한 후 Region of Interest(ROI)를 추출하여 불필요한 영역을 없애고, 왜곡 또한 줄어든다. 본 논문에서 제안하는 기법으로 전처리 후 CNN-LSTM 모델로 성능을 테스트했다. 제안하는 방법으로 전처리를 한 데이터와 하지 않은 데이터로 행동 인식을 한 정확도로 비교하였으며 제안하는 기법으로 전처리 한 데이터로 행동 인식을 한 경우 데이터의 특성에 따라 다르지만, 최대 61%까지 성능향상을 보였다.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.5
• /
• pp.669-674
• /
• 2018

Data augmentation is an efficient way to reduce overfitting on models and to improve a performance supplementing extra data for training. It is more important in deep learning based industrial machine vision. Because deep learning requires huge scale of learning data to learn a model, but acquisition of data can be limited in most of industrial applications. A very generic method for augmenting image data is to perform geometric transformations, such as cropping, rotating, translating and adjusting brightness of the image. The effectiveness of data augmentation in image classification has been reported, but it is rare in defect inspections. We explore and compare various basic augmenting operations for the metal surface defects. The experiments were executed for various types of defects and different CNN networks and analysed for performance improvements by the data augmentations.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.06c
• /
• pp.287-292
• /
• 2007

본 연구에서는 동적 수신호 패턴에 대한 효과적인 인식을 위하여, 특징추출 단계와 패턴 분류 단계의 두 모듈로 이루어지는 복합형 신경망 모델을 제안한다. 특징추출 모듈을 위하여 고유의 특징표현 기법과 3차원 수용영역 구조의 CNN 모델을 제안한다. 이는 3차원 형식의 데이터로 표현되는 수신호 패턴으로부터 특징점의 공간적 변이뿐만 아니라 시간적 변이에 강인한 특징추출 기능을 제공한다. 패턴 분류 모듈에서는 효율적인 학습과 인식 기능을 위하여 수정된 구조의 GFMM 모델을 제안한다. 또한 학습패턴의 빈도를 고려한 활성화 특성과 학습 방법을 정의함으로써 기존의 GFMM 모델이 갖는 단점인 학습결과가 학습순서에 종속되는 특성과 비정상적 패턴 및 노이즈 패턴에 민감한 현상을 개선한다.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.614-615
• /
• 2018

본 연구는 심층학습 기법을 활용하여 양극 데이터에 대해 학습된 모델로부터 예측된 결과를 바탕으로 언어 장애 여부를 판단하고, 이를 바탕으로 효율적인 언어 치료를 수행할 수 있는 방법론을 제시한다. 발화자의 개별 발화에 대해 데이터화를 하여 합성곱 신경망 모델(CNN)을 학습한다. 이를 이용하여 발화자의 연령 집단을 예측하고 결과를 분석하여 발화자의 언어 연령 및 장애 여부를 판단을 할 수 있다.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.522-526
• /
• 2018

본 연구는 수많은 챗봇이 생성될 수 있는 챗봇 빌더 시스템에서 저비용 컴퓨팅 파워에서도 구동 가능한 가벼운 문장 분류 시스템을 제안하며, 미등록어 처리를 위해 워드 임베딩 기법인 GloVe를 이용하여 문장 벡터를 생성하고 이를 추가 자질로 사용하는 방법을 소개한다. 제안한 방법으로 자체 구축한 테스트 말뭉치를 이용하여 성능을 평가해본 결과 최대 93.06% 성능을 보였으며, 자체 보유한 CNN 모델과의 비교 평가 결과 성능은 2.5% 낮지만, 모델 학습 속도는 25배, 학습 시 메모리 사용량은 6배, 생성된 모델 파일 크기는 302배나 효율성 있음을 보였다.

• Journal of Engineering Education Research
• /
• v.23 no.2
• /
• pp.24-30
• /
• 2020

This paper proposes the approaches to the evaluation of learning using concepts of artificial intelligence. Among various techniques, deep learning algorithm is employed to achieve quantitative results of evaluation. In particular, this paper focuses on the process-based evaluation instead of the result-based one using face expression. The expression is simply acquired by digital camera that records face expression when students solve sample test problems. Face expressions are trained using convolutional neural network (CNN) model followed by classification of expression data into three categories, i.e., easy, neutral, difficult. To substantiate the proposed approach, the simulation results show promising results, and this work is expected to open opportunities for intelligent evaluation system in the future.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.398-400
• /
• 2019

Faced with the increasing demand for image recognition on mobile devices, how to run convolutional neural network (CNN) models on mobile devices with limited computing power and limited storage resources encourages people to study efficient model design. In recent years, many effective architectures have been proposed, such as mobilenet_v1, mobilenet_v2 and PeleeNet. However, in the process of feature selection, all these models neglect some information of shallow features, which reduces the capture of shallow feature location and semantics. In this study, we propose an effective framework based on Feature Pyramid Networks to improve the recognition accuracy of deep and shallow images while guaranteeing the recognition speed of PeleeNet structured images. Compared with PeleeNet, the accuracy of structure recognition on CIFA-10 data set increased by 4.0%.