• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권3호
• /
• pp.1188-1203
• /
• 2020

The human expression contains a lot of information that can be used to detect complex conditions such as pain and fatigue. After deep learning became the mainstream method, the traditional feature extraction method no longer has advantages. However, in order to achieve higher accuracy, researchers continue to stack the number of layers of the neural network, which makes the real-time performance of the model weak. Therefore, this paper proposed an expression recognition framework based on densely concatenated convolutional neural networks to balance accuracy and latency and apply it to humanoid robots. The techniques of feature reuse and parameter compression in the framework improved the learning ability of the model and greatly reduced the parameters. Experiments showed that the proposed model can reduce tens of times the parameters at the expense of little accuracy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권12호
• /
• pp.4420-4438
• /
• 2021

Image dehazing is an ill-posed problem which is far from being solved. Traditional image dehazing methods often yield mediocre effects and possess substandard processing speed, while modern deep learning methods perform best only in certain datasets. The haze removal effect when processed by said methods is unsatisfactory, meaning the generalization performance fails to meet the requirements. Concurrently, due to the limited processing speed, most dehazing algorithms cannot be employed in the industry. To alleviate said problems, a lightweight fast dehazing network based on a multiple scale-patch framework (MSP) is proposed in the present paper. Firstly, the multi-scale structure is employed as the backbone network and the multi-patch structure as the supplementary network. Dehazing through a single network causes problems, such as loss of object details and color in some image areas, the multi-patch structure was employed for MSP as an information supplement. In the algorithm image processing module, the image is segmented up and down for processed separately. Secondly, MSP generates a clear dehazing effect and significant robustness when targeting real-world homogeneous and nonhomogeneous hazy maps and different datasets. Compared with existing dehazing methods, MSP demonstrated a fast inference speed and the feasibility of real-time processing. The overall size and model parameters of the entire dehazing model are 20.75M and 6.8M, and the processing time for the single image is 0.026s. Experiments on NTIRE 2018 and NTIRE 2020 demonstrate that MSP can achieve superior performance among the state-of-the-art methods, such as PSNR, SSIM, LPIPS, and individual subjective evaluation.

• 한국인터넷방송통신학회논문지
• /
• 제22권3호
• /
• pp.9-14
• /
• 2022

각종 기기들이 연결되는 사물인터넷(internet of things) 시스템에서 중요한 부품의 고장은 경제적, 인명의 손실을 야기할 수 있다. 시스템 내에서 발생하는 고장으로 인한 손실을 줄이기 위해 고장 검진 기술이 IoT에서 중요한 기술로써 여겨지고 있다. 본 논문에서는 그래프 신경망 기반 방법을 사용하여 시스템 내의 설비에서 취득된 진동 데이터의 특징을 추출하여 고장 여부를 판단하고 유형을 분류하는 방법을 제안한다. 딥러닝 모델의 학습을 위해, CWRU(case western reserve university)에서 취득된 고장 데이터 셋을 입력 데이터로 사용한다. 제안하는 모델의 분류 정확도 성능을 확인하기 위해 기존 제안된 합성곱 신경망(convolutional neural networks) 기반 분류 모델과 제안된 모델을 비교한다. 시뮬레이션 결과, 제안된 모델은 불균등하게 나누어진 데이터에서 기존 모델보다 분류 정확도를 약 5% 향상 시킬 수 있는 것을 확인하였다. 이후 연구로, 제안하는 모델을 경량화해서 분류 속도를 개선할 예정이다.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2007년도 춘계학술대회A
• /
• pp.349-353
• /
• 2007

The main functions of excavator are mainly carried out by excavator attachments such as arm and boom. These components should be designed to be light as well as durable enough because their effects on the whole structure are significant. In this paper, an optimization procedure for lightweight design considering fatigue strength for excavator attachments is presented. The weight of attachments and allowable fatigue stresses at critical areas are used as objective function and constraints, respectively, in which design variables are the thickness of the plates of attachments. The simulated annealing search method is adopted for a global optimization solution. Besides, the response surface method using the artificial neural network is used to simulate constraint function for the sake of practical fast calculation. Some example case of optimization is presented here for a sample excavator. This weight optimization is expected to contribute to a considerable improvement of fuel efficiency of excavator.

• 방송공학회논문지
• /
• 제26권5호
• /
• pp.652-655
• /
• 2021

Versatile Video Coding(VVC)의 압축 효율을 끌어올리기 위하여 다양한 화면 간 예측(inter prediction)기법 중 적응적 움직임 벡터 해상도(Adaptive motion vector resolution, 이하 AMVR)기술이 채택되어 왔다. 다만, AMVR을 적용하여 최적의 해상도를 결정하기 위해서는 매 부호화 유닛마다 다양한 테스트를 진행해야 하며, 이는 율-왜곡 비용의 계산 복잡도 증가를 야기한다. 따라서 VVC의 부호화 복잡도의 감소를 위해 효과적으로 최적의 AMVR 모드를 찾아야 한다. 본 논문에서는 보다 다양한 데이터셋 기반 하에 경량화된 신경망 기반의 AMVR 결정 알고리즘을 제안한다.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2021년도 추계학술대회
• /
• pp.70-73
• /
• 2021

최근 영상 및 비디오 분야에 심층 신경망(DNN, Deep Neural Network)을 사용한 연구가 다양하게 진행됨에 따라 High Dynamic Range (HDR) 이미징 기술에서도 기존의 방법들 보다 우수한 성능을 보이는 심층 신경망 모델들이 등장하였다. 하지만, 심층 신경망을 사용한 방법은 큰 연산량과 많은 GPU 메모리를 사용한다는 문제점이 존재하며, 이는 심층 신경망 기반 기술들의 현실 적용 가능성에 제한이 되고 있다. 이에 본 논문에서는 제한된 연산량과 GPU 메모리 조건에서도 사용 가능한 다중 노출 HDR 경량화 심층 신경망을 제안한다. Kalantari Dataset에 대해 기존 HDR 모델들과의 성능 평가를 진행해 본 결과, PSNR-µ와 PSNR-l 수치에서 GPU 메모리 사용량 대비 우수한 성능을 보임을 확인하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권12호
• /
• pp.3904-3922
• /
• 2022

As a research hotspot, pedestrian detection has a wide range of applications in the field of computer vision in recent years. However, current pedestrian detection methods have problems such as insufficient detection accuracy and large models that are not suitable for large-scale deployment. In view of these problems mentioned above, a lightweight pedestrian detection and early warning method using a new model called you only look once (Yolov5) is proposed in this paper, which utilizing advantages of Yolov5s model to achieve accurate and fast pedestrian recognition. In addition, this paper also optimizes the loss function of the batch normalization (BN) layer. After sparsification, pruning and fine-tuning, got a lot of optimization, the size of the model on the edge of the computing power is lower equipment can be deployed. Finally, from the experimental data presented in this paper, under the training of the road pedestrian dataset that we collected and processed independently, the Yolov5s model has certain advantages in terms of precision and other indicators compared with traditional single shot multiBox detector (SSD) model and fast region-convolutional neural network (Fast R-CNN) model. After pruning and lightweight, the size of training model is greatly reduced without a significant reduction in accuracy, and the final precision reaches 87%, while the model size is reduced to 7,723 KB.

• 대한임베디드공학회논문지
• /
• 제19권3호
• /
• pp.151-158
• /
• 2024

Although on-device artificial intelligence (AI) has gained attention to diagnosing machine faults in real time, most previous studies did not consider the model retraining and redeployment processes that must be performed in real-world industrial environments. Our study addresses this challenge by proposing an on-device AI-based real-time machine fault diagnosis system that utilizes continual learning. Our proposed system includes a lightweight convolutional neural network (CNN) model, a continual learning algorithm, and a real-time monitoring service. First, we developed a lightweight 1D CNN model to reduce the cost of model deployment and enable real-time inference on the target edge device with limited computing resources. We then compared the performance of five continual learning algorithms with three public bearing fault datasets and selected the most effective algorithm for our system. Finally, we implemented a real-time monitoring service using an open-source data visualization framework. In the performance comparison results between continual learning algorithms, we found that the replay-based algorithms outperformed the regularization-based algorithms, and the experience replay (ER) algorithm had the best diagnostic accuracy. We further tuned the number and length of data samples used for a memory buffer of the ER algorithm to maximize its performance. We confirmed that the performance of the ER algorithm becomes higher when a longer data length is used. Consequently, the proposed system showed an accuracy of 98.7%, while only 16.5% of the previous data was stored in memory buffer. Our lightweight CNN model was also able to diagnose a fault type of one data sample within 3.76 ms on the Raspberry Pi 4B device.

• 대한임베디드공학회논문지
• /
• 제17권6호
• /
• pp.367-374
• /
• 2022

The acceleration of neural networks has become an important topic in the field of computer vision. An accelerator is absolutely necessary for accelerating the lightweight model. Most accelerator-supported operators focused on direct convolution operations. If the accelerator does not provide GEMM operation, it is mostly replaced by CPU operation. In this paper, we proposed an optimization technique for 2-stage tiling-based GEMM routines on VTA. We improved performance of the matrix multiplication routine by maximizing the reusability of the input matrix and optimizing the operation pipelining. In addition, we applied the proposed technique to the DarkNet framework to check the performance improvement of the matrix multiplication routine. The proposed GEMM method showed a performance improvement of more than 2.4 times compared to the non-optimized GEMM method. The inference performance of our DarkNet framework has also improved by at least 2.3 times.

• 스마트미디어저널
• /
• 제13권6호
• /
• pp.16-23
• /
• 2024

돼지는 꿀꿀거림, 기침, 비명과 같은 다양한 소리로 환경에 대한 반응과 건강 상태를 나타낸다. 돼지 음성의 중요성으로 최근 들어 돼지의 음성은 축산업 종사자에게 매우 중요한 데이터로 활발하게 연구되고 있다. 이를 위해 돼지의 음성 패턴을 분석하여 농장 소음 속에서 돼지의 음성을 구분하고 음성과 기침 소리를 구분하는 경량화 MobileNet 모델을 제안한다. 이 MobileNet은 돈사 내에서 다양한 배경 잡음, 기침 소리 등의 다양한 소리 속에서 돼지의 음성만을 정밀하게 구분하고 분석할 수 있었다. 테스트 결과, 이 모델은 98.2%의 높은 정확도를 보여주었다. 이러한 결과를 바탕으로 향후 연구에서는 돼지의 감정 분석, 스트레스 파악 등의 문제 해결을 기대한다.