• 대한전자공학회논문지SP
• /
• 제42권4호
• /
• pp.39-48
• /
• 2005

본 논문에서는 Bayesian 추정법과 신경회로망을 이용한 새로운 결 분할 방법을 제안한다 신경회로망의 입력으로는 다중스케일을 가지는 웨이블릿 계수와 인접한 이웃 웨이블릿 계수들의 문맥정보를 사용하고, 신경회로망의 출력을 사후 확률로 모델링한다. 문맥정보는 HMT(Hidden Markov Tree) 모델을 이용하여 구한다. 제안 방법은 HMT를 이용한 ML(Maximum Likelihood) 분할 보다 더 우수한 결과를 보여준다. 또한 HMT를 이용한 결 분할 방법과 제안 방법을 이용한 결 분할 각각에 HMTseg라고 불리는 다중 스케일 Bayesian 영상 분할 기술을 이용하여 후처리를 행한 결 분할 또한 제안 방법이 우수함을 보여준다.

• 한국음향학회지
• /
• 제39권6호
• /
• pp.600-605
• /
• 2020

본 논문에서는 청각 장애인을 위한 소리 감지 홈 모니터링을 위해 다채널 다중 스케일 신경망을 사용한 사운드 이벤트 검출 방식을 제안한다. 제안하는 시스템에서는 홈 내의 여러 무선 마이크 센서들로부터 높은 신호 품질을 갖는 두 개의 채널을 선택하고, 그 신호들로부터 도착신호 지연시간, 피치 범위, 그리고 다중 스케일 합성 곱 신경망을 로그멜 스펙트로그램에 적용하여 추출한 특징들을 양방향 게이트 순환 신경망 기반의 분류기에 적용함으로써 사운드 이벤트 검출의 성능을 더욱 향상시킨다. 검출된 사운드 이벤트 결과는 선택된 채널의 센서 위치와 함께 텍스트로 변환되어 청각 장애인에게 제공된다. 실험결과는 제안한 시스템의 사운드 이벤트 검출 방식이 기존 방식보다 우수하며 청각 장애인에게 효과적으로 사운드 정보를 전달할 수 있음을 보인다.

• Journal of Multimedia Information System
• /
• 제8권4호
• /
• pp.203-210
• /
• 2021

Deep leaning convolutional neural networks (CNN) have successfully been applied to image super-resolution (SR). Despite their great performances, SR techniques tend to focus on a certain upscale factor when training a particular model. Algorithms for single model multi-scale networks can easily be constructed if images are upscaled prior to input, but sub-pixel convolution upsampling works differently for each scale factor. Recent SR methods employ multi-scale and multi-path learning as a solution. However, this causes unshared parameters and unbalanced parameter distribution across various scale factors. We present a multi-scale single-path upsample module as a solution by exploiting the advantages of sub-pixel convolution and interpolation algorithms. The proposed model employs sub-pixel convolution for the highest scale factor among the learning upscale factors, and then utilize 1-dimension interpolation, compressing the learned features on the channel axis to match the desired output image size. Experiments are performed for the single-path upsample module, and compared to the multi-path upsample module. Based on the experimental results, the proposed algorithm reduces the upsample module's parameters by 24% and presents slightly to better performance compared to the previous algorithm.

• International journal of advanced smart convergence
• /
• 제7권2호
• /
• pp.95-100
• /
• 2018

In this paper, we propose a binarized multi-scale module to accelerate the speed of the pose estimating deep neural network. Recently, deep learning is also used for fine-tuned tasks such as pose estimation. One of the best performing pose estimation methods is based on the usage of two neural networks where one computes the heat maps of the body parts and the other computes the part affinity fields between the body parts. However, the convolution filtering with a large kernel filter takes much time in this model. To accelerate the speed in this model, we propose to change the large kernel filters with binarized multi-scale modules. The large receptive field is captured by the multi-scale structure which also prevents the dropdown of the accuracy in the binarized module. The computation cost and number of parameters becomes small which results in increased speed performance.

• 한국멀티미디어학회논문지
• /
• 제21권12호
• /
• pp.1467-1472
• /
• 2018

For the problem that the face image of surveillance video cannot be accurately identified due to the low resolution, this paper proposes a low resolution face recognition solution based on convolutional neural network model. Convolutional Neural Networks (CNN) model for multi-scale input The CNN model for multi-scale input is an improvement over the existing "two-step method" in which low-resolution images are up-sampled using a simple bi-cubic interpolation method. Then, the up sampled image and the high-resolution image are mixed as a model training sample. The CNN model learns the common feature space of the high- and low-resolution images, and then measures the feature similarity through the cosine distance. Finally, the recognition result is given. The experiments on the CMU PIE and Extended Yale B datasets show that the accuracy of the model is better than other comparison methods. Compared with the CMDA_BGE algorithm with the highest recognition rate, the accuracy rate is 2.5%~9.9%.

• Wind and Structures
• /
• 제2권1호
• /
• pp.25-40
• /
• 1999

In this paper, artificial neural networks, a new kind of intelligent method, are employed to model and predict amplitude dependent damping in buildings based on our full-scale measurements of buildings. The modelling method and procedure using neural networks to model the damping are studied. Comparative analysis of different neural network models of damping, which includes multi-layer perception network (MLP), recurrent neural network, and general regression neural network (GRNN), is performed and discussed in detail. The performances of the models are evaluated and discussed by tests and predictions including self-test, "one-lag" prediction and "multi-lag" prediction of the damping values at high amplitude levels. The established models of damping are used to predict the damping in the following three ways : (1) the model is established by part of the data measured from one building and is used to predict the another part of damping values which are always difficult to obtain from field measurements : the values at the high amplitude level. (2) The model is established by the damping data measured from one building and is used to predict the variation curve of damping for another building. And (3) the model is established by the data measured from more than one buildings and is used to predict the variation curve of damping for another building. The prediction results are discussed.

• 제어로봇시스템학회논문지
• /
• 제6권12호
• /
• pp.1113-1119
• /
• 2000

To operate a process plant safely and economically, process monitoring is very important. Process monitoring is the task to identify the state of the system from sensor data. Process monitoring includes data acquisition, regulatory control, data reconciliation, fault detection, etc. This research focuses on the data recon-ciliation using scale-space filtering and fault detection using functional-link associative neural networks. Scale-space filtering is a multi-resolution signal analysis method. Scale-space filtering can extract highest frequency factors(noise) effectively. But scale-space filtering has too large calculation costs and end effect problems. This research reduces the calculation cost of scale-space filtering by applying the minimum limit to the gaussian kernel. And the end-effect that occurs at the end of the signal of the scale-space filtering is overcome by using extrapolation related with the clustering change detection method. Nonlinear principal component analysis methods using neural network have been reviewed and the separately expanded functional-link associative neural network is proposed for chemical process monitoring. The separately expanded functional-link associative neural network has better learning capabilities, generalization abilities and short learning time than the exiting-neural networks. Separately expanded functional-link associative neural network can express a statistical model similar to real process by expanding the input data separately. Combining the proposed methods-modified scale-space filtering and fault detection method using the separately expanded functional-link associative neural network-a process monitoring system is proposed in this research. the usefulness of the proposed method is proven by its application a boiler water supply unit.

• 한국인공지능학회지
• /
• 제11권4호
• /
• pp.1-8
• /
• 2023

In this paper, we propose a black ice detection platform framework using Convolutional Neural Networks (CNNs). To overcome black ice problem, we introduce a real-time based early warning platform using CNN-based architecture, and furthermore, in order to enhance the accuracy of black ice detection, we apply a multi-scale dilation convolution feature fusion (MsDC-FF) technique. Then, we establish a specialized experimental platform by using a comprehensive dataset of thermal road black ice images for a training and evaluation purpose. Experimental results of a real-time black ice detection platform show the better performance of our proposed network model compared to conventional image segmentation models. Our proposed platform have achieved real-time segmentation of road black ice areas by deploying a road black ice area segmentation network on the edge device Jetson Nano devices. This approach in parallel using multi-scale dilated convolutions with different dilation rates had faster segmentation speeds due to its smaller model parameters. The proposed MsCD-FF Net(2) model had the fastest segmentation speed at 5.53 frame per second (FPS). Thereby encouraging safe driving for motorists and providing decision support for road surface management in the road traffic monitoring department.

• 한국전자통신학회논문지
• /
• 제19권1호
• /
• pp.301-307
• /
• 2024

본 연구는 다양한 스케일 조건에서 열화상 이미지를 향상시키기 위한 새로운 캐스케이드 융합 구조를 제안한다. 특정 스케일에 맞춰 설계된 방법들은 다중 스케일에서 열화상 이미지 처리에 한계가 있었다. 이를 극복하기 위해 본 논문에서는 다중 스케일 표현을 활용하는 캐스케이드 특징 융합 기법에 기반한 통합 프레임워크를 제시한다. 서로 다른 스케일의 신뢰도 맵을 순차적으로 융합함으로써 스케일에 제약받지 않는 학습이 가능해진다. 제안된 구조는 상호 스케일 의존성을 강화하기 위해 엔드 투 엔드 방식으로 훈련된 합성곱 신경망으로 구성되어 있다. 실험 결과, 제안된 방법은 기존의 다중 스케일 열화상 이미지 향상 방법들보다 우수한 성능을 보인다는 것을 확인할 수 있었다. 또한, 실험 데이터셋에 대한 성능 분석 결과 이미지 품질 지표가 일관되게 개선되었으며, 이는 캐스케이드 융합 설계가 스케일 간 견고한 일반화를 가능하게 하고 교차 스케일 표현 학습을 더 효율적으로 수행하는 데 기여하는 것을 보여준다.

• Journal of Multimedia Information System
• /
• 제8권2호
• /
• pp.101-110
• /
• 2021

In computer vision, single-image super resolution has been an area of research for a significant period. Traditional techniques involve interpolation-based methods such as Nearest-neighbor, Bilinear, and Bicubic for image restoration. Although implementations of convolutional neural networks have provided outstanding results in recent years, efficiency and single model multi-scalability have been its challenges. Furthermore, previous works haven't placed enough emphasis on real-number scalability. Interpolation-based techniques, however, have no limit in terms of scalability as they are able to upscale images to any desired size. In this paper, we propose a convolutional neural network possessing the advantages of the interpolation-based techniques, which is also efficient, deeming it suitable in practical implementations. It consists of convolutional layers applied on the low-resolution space, post-up-sampling along the end hidden layers, and additional layers on high-resolution space. Up-sampling is applied on a multiple channeled feature map via bicubic interpolation using a single model. Experiments on architectural structure, layer reduction, and real-number scale training are executed with results proving efficient amongst multi-scale learning (including scale multi-path-learning) based models.