• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제9권12호
• /
• pp.419-430
• /
• 2020

본 연구는 기존 전력량 예측 모델의 구조를 변경하여 모델의 예측 능력을 향상 시킬 수 있는 방법에 관하여 연구하였다. 전기에 대한 수요는 그 어느 때보다 증가하고 있다. 산업 부문에서는 그 어느 부문 보다 전기 소모량이 많음으로, 더욱 정확한 공장 지역의 전력량 소모 예측 모델이 잉여 에너지 생산을 줄이기 위해 주목을 받고 있다. 우리는 2개의 개별 encoder와 한개의 decoder를 사용하여, 장기와 단기 데이터를 모두 사용하는 double encoder-decoder 모델을 제안한다. 우리는 제안된 모델을 세홍(주)의 생산 구역에서 2019년 1월 1일부터 2019년 6월 30일 까지 모집된 전력 소모량 데이터에서 평가 하였다. double encoder-decoder 모델은 기존의 encoder-decoder 모델을 사용했을 때와 비교하여 약 10 %의 평균 절대 비율 오차의 감소를 기록 하였다. 본 결과는 제안한 모델이 encoder-decoder 모델에 비해 생산 지역의 전력 사용량의 예측을 더 정확하게 하는 모델임을 보여준다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권9호
• /
• pp.3823-3840
• /
• 2020

Recently, auto-encoder has emerged as the most popular method in convolutional neural network (CNN) based image compression and has achieved impressive performance. In the traditional auto-encoder based image compression model, the encoder simply sends the features of last layer to the decoder, which cannot allocate bits over different spatial regions in an efficient way. Besides, these methods do not fully exploit the contextual information under different receptive fields for better reconstruction performance. In this paper, to solve these issues, a novel auto-encoder model is designed for image compression, which can effectively transmit the hierarchical features of the encoder to the decoder. Specifically, we first propose an adaptive bit-allocation strategy, which can adaptively select an importance channel. Then, we conduct the multiply operation on the generated importance mask and the features of the last layer in our proposed encoder to achieve efficient bit allocation. Moreover, we present an additional novel perceptual loss function for more accurate image details. Extensive experiments demonstrated that the proposed model can achieve significant superiority compared with JPEG and JPEG2000 both in both subjective and objective quality. Besides, our model shows better performance than the state-of-the-art convolutional neural network (CNN)-based image compression methods in terms of PSNR.

• International Journal of Precision Engineering and Manufacturing
• /
• 제8권2호
• /
• pp.20-25
• /
• 2007

This paper describes a second-generation dual sine-grid surface encoder for 2-D position measurements. The surface encoder consisted of a 2-D grid with a 2-D sinusoidal pattern on its surface, and a 2-D angle sensor that detected the 2-D profile of the surface grid The 2-D angle sensor design of previously developed first-generation surface encoders was based on geometric optics. To improve the resolution of the surface encoder, we fabricated a 2-D sine-grid with a pitch of $10{\mu}m$. We also established a new optical model for the second-generation surface encoder that utilizes diffraction and interference to generate its measured values. The 2-D sine-grid was fabricated on a workpiece by an ultra precision lathe with the assistance of a fast tool servo. We then performed a UV-casting process to imprint the sine-grid on a transparent plastic film and constructed an experimental setup to realize the second-generation surface encoder. We conducted tests that demonstrated the feasibility of the proposed surface encoder model.

• 인터넷정보학회논문지
• /
• 제22권2호
• /
• pp.1-9
• /
• 2021

전통적으로 대부분의 악성코드는 도메인 전문가에 의해 추출된 특징 정보를 활용하여 분석되었다. 하지만 이러한 특징 기반의 분석방식은 분석가의 역량에 의존적이며 기존의 악성코드를 변형한 변종 악성코드를 탐지하는 데 한계를 가지고 있다. 본 연구에서는 도메인 전문가의 개입 없이도 변종 악성코드의 패밀리를 분류할 수 있는 ResNet-Variational AutoEncder 기반 변종 악성코드 분류 방법을 제안한다. Variational AutoEncoder 네트워크는 입력값으로 제공되는 훈련 데이터의 학습 과정에서 데이터의 특징을 잘 이해하며 정규 분포 내에서 새로운 데이터를 생성하는 특징을 가지고 있다. 본 연구에서는 Variational AutoEncoder의 학습 과정에서 잠재 변수를 추출을 통해 악성코드의 중요 특징을 추출할 수 있었다. 또한 훈련 데이터의 특징을 더욱 잘 학습하고 학습의 효율성을 높이기 위해 전이 학습을 수행했다. ImageNet Dataset으로 사전학습된 ResNet-152 모델의 학습 파라미터를 Encoder Network의 학습 파라미터로 전이했다. 전이학습을 수행한 ResNet-Variational AutoEncoder의 경우 기존 Variational AutoEncoder에 비해 높은 성능을 보였으며 학습의 효율성을 제공하였다. 한편 변종 악성코드 분류를 위한 방법으로는 앙상블 모델인 Stacking Classifier가 사용되었다. ResNet-VAE 모델의 Encoder Network로 추출한 변종 악성코드 특징 데이터를 바탕으로 Stacking Classifier를 학습한 결과 98.66%의 Accuracy와 98.68의 F1-Score를 얻을 수 있었다.

• Journal of Positioning, Navigation, and Timing
• /
• 제8권4호
• /
• pp.193-200
• /
• 2019

The auto-encoder network which is a good candidate to handle the modeling of the signal strength attenuation is designed for denoising and compensating the distortion of the received data. It provides a non-linear mapping function by iteratively learning the encoder and the decoder. The encoder is the non-linear mapping function, and the decoder demands accurate data reconstruction from the representation generated by the encoder. In addition, the adaptive network width which supports the automatic generation of new hidden nodes and pruning of inconsequential nodes is also implemented in the proposed algorithm for increasing the efficiency of the algorithm. Simulation results show that the proposed method can improve the neural network training surface to achieve the highest possible accuracy of the signal modeling compared with the conventional modeling method.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 Ⅳ
• /
• pp.2467-2470
• /
• 2003

This paper presents a new structure of MPEG-4 AAC encoder. The proposed encoder directly shapes quantization noise distribution according to the energy distribution curve and thereafter performs adjustment of the offset level of the noise distribution to meet the given bit rate. The direct noise shaping and the bit rate matching scheme of the proposed encoder algorithm significantly alleviate the problem of conventional encoder's processing burden related with the employment of the precise psychoacoustic model and iteration intensive quantizer. The encoder algorithm is implemented on ARM processor with fixed-feint arithmetic operations. The audio quality of the implemented system is observed comparable to those of commercially available encoders, white the complexity of the implementation is drastically reduced in comparison to the conventional encoder systems.

• 대한임베디드공학회논문지
• /
• 제9권2호
• /
• pp.117-125
• /
• 2014

The EtherCAT protocol has become a very promising alternative for real-time industrial application due to its ability to eliminate uncertainties in the Ethernet. However, the extended markup language (XML) for the EtherCAT network system, which is required in the design, lacks systematic development to take advantage of model transformation techniques. This paper focuses on the system development procedure of the EtherCAT-based encoder system using the CANopen over EtherCAT (CoE) protocol. UML modeling is being adapted to design for XML of EtherCAT-based encoder system. To this purpose, this paper analyzes the object dictionary (OD) of a commercial encoder and CANopen over EtherCAT. A UML diagram is then designed based on the analysis, and XML is generated through the designed UML diagram. Finally, an experimental test_bed for the EtherCAT-based encoder system is implemented and its performance is compared with a commercial encoder.

• 정보과학회 논문지
• /
• 제45권2호
• /
• pp.157-164
• /
• 2018

딥러닝의 급속한 발전은 패턴인식 분야의 성능을 혁신했으며 몇몇 문제에서는 인간 수준을 넘어서는 결과들을 보여주고 있다. 데이타를 분류하는 패턴인식과 달리 본 논문에서는 주어진 몇개의 한국어 문장으로부터 비슷한 문장들을 생성하는 문제를 다룬다. 이를위해 생성모델 중의 하나인 Variational Auto-Encoder 기반의 모델을 한국어 생성에 맞게 개선하고 적용하는 방법들을 논의한다. 첫째, 교착어인 한국어의 특성상 띄어쓰기를 기준으로 단어 생성시 단어의 개수가 너무 많아 이를 줄이기 위해 조사 및 어미들을 분리할 필요가 있다. 둘째, 한국어는 어순이 비교적 자유롭고 주어 목적어 등이 생략되는 경우가 많아 기존의 단방향 인코더를 양방향으로 확장한다. 마지막으로, 주어진 문장들을 기반으로 비슷하지만 새로운 문장들을 생성하기 위해 기존 문장들의 인코딩된 벡터표현들로부터 새로운 벡터를 찾아내고, 이 벡터를 디코딩하여 문장을 생성한다. 실험 결과를 통해 제안한 방법의 성능을 확인한다.

• 로봇학회논문지
• /
• 제18권4호
• /
• pp.427-435
• /
• 2023

This paper proposes a method of estimating the pose of a mobile robot by using a learning model. When estimating the pose of a mobile robot, wheel encoder and inertial measurement unit (IMU) data are generally utilized. However, depending on the condition of the ground surface, slip occurs due to interaction between the wheel and the floor. In this case, it is hard to predict pose accurately by using only encoder and IMU. Thus, in order to reduce pose error even in such conditions, this paper introduces a pose estimation method based on a learning model using data of the wheel encoder and IMU. As the learning model, long short-term memory (LSTM) network is adopted. The inputs to LSTM are velocity and acceleration data from the wheel encoder and IMU. Outputs from network are corrected linear and angular velocity. Estimated pose is calculated through numerically integrating output velocities. Dataset used as ground truth of learning model is collected in various ground conditions. Experimental results demonstrate that proposed learning model has higher accuracy of pose estimation than extended Kalman filter (EKF) and other learning models using the same data under various ground conditions.

• International journal of advanced smart convergence
• /
• 제12권4호
• /
• pp.142-146
• /
• 2023

We present a method for generating 3D structures and rendering objects by combining VAE (Variational Autoencoder) and GAN (Generative Adversarial Network). This approach focuses on generating and rendering 3D models with improved quality using residual learning as the learning method for the encoder. We deep stack the encoder layers to accurately reflect the features of the image and apply residual blocks to solve the problems of deep layers to improve the encoder performance. This solves the problems of gradient vanishing and exploding, which are problems when constructing a deep neural network, and creates a 3D model of improved quality. To accurately extract image features, we construct deep layers of the encoder model and apply the residual function to learning to model with more detailed information. The generated model has more detailed voxels for more accurate representation, is rendered by adding materials and lighting, and is finally converted into a mesh model. 3D models have excellent visual quality and accuracy, making them useful in various fields such as virtual reality, game development, and metaverse.