• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.59-68
• /
• 2021

As artificial intelligence(AI) technologies, which have made rapid growth recently, began to be applied to the marine environment such as ships, there have been active researches on the application of CNN-based models specialized for digital videos. In E-Navigation service, which is combined with various technologies to detect floating objects of clash risk to reduce human errors and prevent fires inside ships, real-time processing is of huge importance. More functions added, however, mean a need for high-performance processes, which raises prices and poses a cost burden on shipowners. This study thus set out to propose a method capable of processing information at a high rate while maintaining the accuracy by applying Quantization techniques of a deep learning model. First, videos were pre-processed fit for the detection of floating matters in the sea to ensure the efficient transmission of video data to the deep learning entry. Secondly, the quantization technique, one of lightweight techniques for a deep learning model, was applied to reduce the usage rate of memory and increase the processing speed. Finally, the proposed deep learning model to which video pre-processing and quantization were applied was applied to various embedded boards to measure its accuracy and processing speed and test its performance. The proposed method was able to reduce the usage of memory capacity four times and improve the processing speed about four to five times while maintaining the old accuracy of recognition.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.4
• /
• pp.41-46
• /
• 2023

Human urine is a process of excreting waste products in the blood, and it is easy to collect and contains various substances. Urinalysis is used to check for diseases, health conditions, and urinary tract infections. There are three methods of urinalysis: physical property test, chemical test, and microscopic test, and chemical test results can be easily confirmed using urine test strips. A variety of items can be tested on the urine test strip, through which various diseases can be identified. Recently, with the spread of smart phones, research on reading urine test strips using smart phones is being conducted. There is a method of detecting and reading the color change of a urine test strip using a smartphone. This method uses the RGB values and the color difference formula to discriminate. However, there is a problem in that accuracy is lowered due to various environmental factors. This paper applies a deep learning model to solve this problem. In particular, color discrimination of a urine test strip is improved in a smartphone using a lightweight CNN (Convolutional Neural Networks) model. CNN is a useful model for image recognition and pattern finding, and a lightweight version is also available. Through this, it is possible to operate a deep learning model on a smartphone and extract accurate urine test results. Urine test strips were taken in various environments to prepare deep learning model training images, and a urine test service application was designed using MobileNet V3.

• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.167-174
• /
• 2021

Recently, deep learning technology is widely used in various computer vision applications, such as object recognition, classification, and image generation. In particular, the deep learning-based super-resolution has been gaining significant performance improvement. Fast super-resolution convolutional neural network (FSRCNN) is a well-known model as a deep learning-based super-resolution algorithm that output image is generated by a deconvolutional layer. In this paper, we propose an FPGA-based convolutional neural networks accelerator that considers parallel computing efficiency. In addition, the proposed method proposes Optimal-FSRCNN, which is modified the structure of FSRCNN. The number of multipliers is compressed by 3.47 times compared to FSRCNN. Moreover, PSNR has similar performance to FSRCNN. We developed a real-time image processing technology that implements on FPGA.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.4
• /
• pp.95-105
• /
• 2021

Road surface damage detection is essential for a comfortable driving environment and the prevention of safety accidents. Road management institutes are using automated technology-based inspection equipment and systems. As one of these automation technologies, a sensor to detect road surface damage plays an important role. For this purpose, several studies on sensors using deep learning have been conducted in recent years. Road images and label images are needed to develop such deep learning algorithms. On the other hand, considerable time and labor will be needed to secure label images. In this paper, the adversarial learning method, one of the semi-supervised learning techniques, was proposed to solve this problem. For its implementation, a lightweight deep neural network model was trained using 5,327 road images and 1,327 label images. After experimenting with 400 road images, a model with a mean intersection over a union of 80.54% and an F1 score of 77.85% was developed. Through this, a technology that can improve recognition performance by adding only road images was developed to learning without label images and is expected to be used as a technology for road surface management in the future.

• Smart Media Journal
• /
• v.12 no.5
• /
• pp.28-35
• /
• 2023

In this study, we propose a new light-weight model RoutingConvNet with fewer parameters to improve the applicability and practicality of speech emotion recognition. To reduce the number of learnable parameters, the proposed model connects bidirectional MFCCs on a channel-by-channel basis to learn long-term emotion dependence and extract contextual features. A light-weight deep CNN is constructed for low-level feature extraction, and self-attention is used to obtain information about channel and spatial signals in speech signals. In addition, we apply dynamic routing to improve the accuracy and construct a model that is robust to feature variations. The proposed model shows parameter reduction and accuracy improvement in the overall experiments of speech emotion datasets (EMO-DB, RAVDESS, and IEMOCAP), achieving 87.86%, 83.44%, and 66.06% accuracy respectively with about 156,000 parameters. In this study, we proposed a metric to calculate the trade-off between the number of parameters and accuracy for performance evaluation against light-weight.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.1
• /
• pp.21-26
• /
• 2023

Knowledge distillation (KD) is a model lightening technology that transfers the knowledge of deep models to light models. Most KD methods have been developed for classification models, and there have been few KD studies in the field of super-resolution (SR). In this paper, various KD methods are applied to an SR model and their performance is compared. Specifically, we modified the loss function to apply each KD method to the SR model and conducted an experiment to learn a student model that was about 27 times lighter than the teacher model and to double the image resolution. Through the experiment, it was confirmed that some KD methods were not valid when applied to SR models, and that the performance was the highest when the relational KD and the traditional KD methods were combined.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.604-606
• /
• 2024

컴퓨터 비전 응용은 우리 생활에서 중요한 역할을 한다. 현재, 대규모 모델의 등장으로 딥 러닝의 훈련 및 운행 비용이 급격히 상승하고 있다. 자원이 제한된 환경에서는 일부 AI 프로그램을 실행할 수 없게 되므로, 경량화 연구가 필요하다. YOLOv8 은 현재 주요 목표 검출 모델 중 하나이며, 본 논문은 다중 스케일 Ghost 컨볼루션 모듈을 사용하여 구축된 새로운 YOLOv8-pose-msg 키포인트 검출 모델을 제안한다. 다양한 사양에서 새 모델의 매개변수 양은 최소 34% 감소할 수 있으며, 최대 59%까지 감소할 수 있다. 종합적인 검출 성능은 비교적 대규모 데이터셋에서 원래의 수준을 유지할 수 있으며, 소규모 데이터셋에서의 키포인트 검출은 30% 이상 증가할 수 있다. 동시에 최대 25%의 훈련 및 추론 시간을 절약할 수 있다.

• Journal of Convergence for Information Technology
• /
• v.10 no.8
• /
• pp.23-34
• /
• 2020

Recently, the artificial intelligence deep learning field has been hard to commercialize due to the high computing power and the price problem of computing resources. In this paper, we apply a double pruning techniques to evaluate the performance of the in-depth neural network and various datasets. Double pruning combines basic Network-slimming and Parameter-prunning. Our proposed technique has the advantage of reducing the parameters that are not important to the existing learning and improving the speed without compromising the learning accuracy. After training various datasets, the pruning ratio was increased to reduce the size of the model.We confirmed that MobileNet-V3 showed the highest performance as a result of NetScore performance analysis. We confirmed that the performance after pruning was the highest in MobileNet-V3 consisting of depthwise seperable convolution neural networks in the Cifar 10 dataset, and VGGNet and ResNet in traditional convolutional neural networks also increased significantly.

• Journal of the Korea Convergence Society
• /
• v.12 no.12
• /
• pp.31-37
• /
• 2021

In this paper proposed a feature fusion and spatial attention-based modified YOLOv4S for small and occluded detection. Conventional YOLOv4S is a lightweight network and lacks feature extraction capability compared to the method of the deep network. The proposed method first combines feature maps of different scales with feature fusion to enhance semantic and low-level information. In addition expanding the receptive field with dilated convolution, the detection accuracy for small and occluded objects was improved. Second by improving the conventional spatial information with spatial attention, the detection accuracy of objects classified and occluded between objects was improved. PASCAL VOC and COCO datasets were used for quantitative evaluation of the proposed method. The proposed method improved mAP by 2.7% in the PASCAL VOC dataset and 1.8% in the COCO dataset compared to the Conventional YOLOv4S.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.9
• /
• pp.1330-1339
• /
• 2022

3D object detection generally aims to detect relatively large data such as automobiles, buses, persons, furniture, etc, so it is vulnerable to small object detection. In addition, in an environment with limited resources such as embedded devices, it is difficult to apply the model because of the huge amount of computation. In this paper, the accuracy of small object detection was improved by focusing on local features using only one layer, and the inference speed was improved through the proposed knowledge distillation method from large pre-trained network to small network and adaptive quantization method according to the parameter size. The proposed model was evaluated using SUN RGB-D Val and self-made apple tree data set. Finally, it achieved the accuracy performance of 62.04% at mAP@0.25 and 47.1% at mAP@0.5, and the inference speed was 120.5 scenes per sec, showing a fast real-time processing speed.