• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.2
• /
• pp.84-90
• /
• 2022

Machine tool state monitoring is a process that automatically detects the states of machine. In the manufacturing process, the efficiency of machining and the quality of the product are affected by the condition of the tool. Wear and broken tools can cause more serious problems in process performance and lower product quality. Therefore, it is necessary to develop a system to prevent tool wear and damage during the process so that the tool can be replaced in a timely manner. This paper proposes a method for diagnosing five tool states using a deep learning-based hierarchical convolutional neural network to change tools at the right time. The one-dimensional acoustic signal generated when the machine cuts the workpiece is converted into a frequency-based power spectral density two-dimensional image and use as an input for a convolutional neural network. The learning model diagnoses five tool states through three hierarchical steps. The proposed method showed high accuracy compared to the conventional method. In addition, it will be able to be utilized in a smart factory fault diagnosis system that can monitor various machine tools through real-time connecting.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.4
• /
• pp.234-240
• /
• 2022

This paper proposes a customized AI exercise recommendation service for balancing the relative amount of exercise according to the working environment by each occupation. WISDM database is collected by using acceleration and gyro sensors, and is a dataset that classifies physical activities into 18 categories. Our system recommends a adaptive exercise using the analyzed activity type after classifying 18 physical activities into 3 physical activities types such as whole body, upper body and lower body. 1 Dimensional convolutional neural network is used for classifying a physical activity in this paper. Proposed model is composed of a convolution blocks in which 1D convolution layers with a various sized kernel are connected in parallel. Convolution blocks can extract a detailed local features of input pattern effectively that can be extracted from deep neural network models, as applying multi 1D convolution layers to input pattern. To evaluate performance of the proposed neural network model, as a result of comparing the previous recurrent neural network, our method showed a remarkable 98.4% accuracy.

• KIISE Transactions on Computing Practices
• /
• v.22 no.11
• /
• pp.614-618
• /
• 2016

As information on the internet and the data from smart devices are growing, the amount of stream data is also increasing in the real world. The stream data, which is a potentially large data, requires online learnable models and algorithms. In this paper, we propose a novel class of models: predictive convolutional neural networks to be able to perform online learning. These models are designed to deal with longer patterns as the layers become higher due to layering convolutional operations: detection and max-pooling on the time axis. As a preliminary check of the concept, we chose two-month gathered GPS data sequence as an observation sequence. On learning them with the proposed method, we compared the original sequence and the regenerated sequence from the abstract information of the models. The result shows that the models can encode long-range patterns, and can generate a raw observation sequence within a low error.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.4
• /
• pp.161-166
• /
• 2018

In this paper, we propose a method for the random selection of pooling operations for the regularization and reduction of cross validation in convolutional neural networks. The pooling operation in convolutional neural networks is used to reduce the size of the feature map and for its shift invariant properties. In the existing pooling method, one pooling operation is applied in each pooling layer. Because this method fixes the convolution network, the network suffers from overfitting, which means that it excessively fits the models to the training samples. In addition, to find the best combination of pooling operations to maximize the performance, cross validation must be performed. To solve these problems, we introduce the probability concept into the pooling layers. The proposed method does not select one pooling operation in each pooling layer. Instead, we randomly select one pooling operation among multiple pooling operations in each pooling region during training, and for testing purposes, we use probabilistic weighting to produce the expected output. The proposed method can be seen as a technique in which many networks are approximately averaged using a different pooling operation in each pooling region. Therefore, this method avoids the overfitting problem, as well as reducing the amount of cross validation. The experimental results show that the proposed method can achieve better generalization performance and reduce the need for cross validation.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.659-660
• /
• 2021

본 논문에서는 데카르트 좌표계 기반으로 노드를 압축함으로써 SR(Super-resolution) 기반 연기 합성을 효율적으로 처리할 수 있는 방법을 제안한다. 제안하는 방법은 다운 스케일링과 이진화를 통하여 연기 시뮬레이션의 계산 공간을 효율적으로 줄이고, 데카르트 좌표계 축을 기준으로 쿼드트리의 말단 노드를 압축함으로써 네트워크의 입력으로 전달하는 데이터 개수를 줄인다. 학습에 사용된 데이터는 COCO 2017 데이터셋이며, 인공신경망은 VGG19 기반 네트워크를 사용한다. 컨볼루션 계층을 거칠 때 데이터의 손실을 막기 위해 잔차(Residual)방식과 유사하게 이전 계층의 출력 값을 더해주며 학습한다. 결과적으로 제안하는 방법은 이전 결과에 비해 네트워크로 전달해야 하는 데이터가 압축되어 개수가 줄어드는 결과를 얻었으며, 그로 인해 네트워크 단계에서 필요한 I/O 과정을 효율적으로 처리할 수 있게 되었다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.01a
• /
• pp.261-264
• /
• 2021

본 논문에서는 SR(Super-Resolution)을 계산하는데 필요한 데이터를 효율적으로 분류하고 분할하여 빠르게 SR연산을 가능하게 하는 쿼드트리 기반 최적화 기법을 제안한다. 제안하는 방법은 입력 데이터로 사용하는 연기 데이터를 다운스케일링(Downscaling)하여 쿼드트리 연산 소요 시간을 감소시키며, 이때 연기의 밀도를 이진화함으로써, 다운스케일링 과정에서 밀도가 손실되는 문제를 피한다. 학습에 사용된 데이터는 COCO 2017 Dataset이며, 인공신경망은 VGG19 기반 네트워크를 사용한다. 컨볼루션 계층을 거칠 때 데이터의 손실을 막기 위해 잔차(Residual)방식과 유사하게 이전 계층의 출력 값을 더해주며 학습한다. 결과적으로 제안하는 방법은 이전 결과 기법에 비해 약15~18배 정도의 속도향상을 얻었다.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.6
• /
• pp.861-868
• /
• 2023

One of the key technologies in providing data analysis in the deep learning while maintaining security is fully homomorphic encryption. Due to constraints in operations on fully homomorphically encrypted data, non-arithmetic functions used in deep learning must be approximated by polynomials. Until now, the degrees of approximation polynomials with composite minimax polynomials have been uniformly set across layers, which poses challenges for effective network designs on fully homomorphic encryption. This study theoretically proves that setting different degrees of approximation polynomials constructed by composite minimax polynomial in each layer does not pose any issues in the inference on convolutional neural networks.

• Journal of KIISE
• /
• v.44 no.12
• /
• pp.1290-1295
• /
• 2017

Document classification, a task of classifying the category of each document based on text, is one of the fundamental areas for natural language processing. Document classification may be used in various fields such as topic classification and sentiment classification. Neural network models for document classification can be divided into two categories: word-level models and character-level models that treat words and characters as basic units respectively. In this study, we propose a neural network model that combines character-level and word-level models to improve performance of document classification. The proposed model extracts the feature vector of each word by combining information obtained from a word embedding matrix and information encoded by a character-level neural network. Based on feature vectors of words, the model classifies documents with a hierarchical structure wherein recurrent neural networks with attention mechanisms are used for both the word and the sentence levels. Experiments on real life datasets demonstrate effectiveness of our proposed model.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.07a
• /
• pp.22-25
• /
• 2020

본 논문에서는 큰 크기의 심층 신경망을 압축하기위해 네트워크 수준의 가중치 공유방법인 Global Weight 패러다임을 최초로 제시한다. 기존의 가중치 공유방법은 계층별로 가중치를 공유하는 것이 대부분이었다. Global Weight 는 기존 방법과 달리 전체 네트워크에서 가중치를 공유하는 효율적인 방법이다. 우리는 Global Weight 를 사용하여 학습되는 새로운 컨볼루션 연산인 Global Weight Convolution(GWConv)연산과 GWConv를 적용한 Global Weight Networks(GWNet)을 제안한다. CIFAR10 데이터셋에서 실험한 결과 2.18 배 압축에서 85.64%, 3.41 배 압축에서 85.46%의 정확도를 보였다. Global Weight 패러다임은 가중치 공유가 궁극적으로 풀고자 했던 중복되는 가중치를 최소화하는 획기적인 방법이며, 추후 심도 있는 연구가 수행될 수 있음을 시사한다.

• Smart Media Journal
• /
• v.5 no.3
• /
• pp.35-41
• /
• 2016

In this paper, we propose hand-gesture signal recognition based on EPS(Electronic Potential Sensor) using Deep learning model. Extracted signals which from Electronic field based sensor, EPS have much of the noise, so it must remove in pre-processing. After the noise are removed with filter using frequency feature, the signals are reconstructed with dimensional transformation to overcome limit which have just one-dimension feature with voltage value for using convolution operation. Then, the reconstructed signal data is finally classified and recognized using multiple learning layers model based on deep learning. Since the statistical model based on probability is sensitive to initial parameters, the result can change after training in modeling phase. Deep learning model can overcome this problem because of several layers in training phase. In experiment, we used two different deep learning structures, Convolutional neural networks and Recurrent Neural Network and compared with statistical model algorithm with four kinds of gestures. The recognition result of method using convolutional neural network is better than other algorithms in EPS gesture signal recognition.