• Journal of Information Processing Systems
• /
• 제18권5호
• /
• pp.701-710
• /
• 2022

With the development of artificial intelligence technology, various methods have been developed for recognizing objects in images using machine learning. Image segmentation is the most effective among these methods for recognizing objects within an image. Conventionally, image datasets of various classes are trained simultaneously. In situations where several classes require segmentation, all datasets have to be trained thoroughly. Such repeated training results in low training efficiency because most of the classes have already been trained. In addition, the number of classes that appear in the datasets affects training. Some classes appear in datasets in remarkably smaller numbers than others, and hence, the training errors will not be properly reflected when all the classes are trained simultaneously. Therefore, a new method that separates some classes from the dataset is proposed to improve efficiency during training. In addition, the accuracies of the conventional and proposed methods are compared.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2019년도 제59차 동계학술대회논문집 27권1호
• /
• pp.123-124
• /
• 2019

This paper proposes an improved deep learning method based on small data sets for animal image classification. Firstly, we use a CNN to build a training model for small data sets, and use data augmentation to expand the data samples of the training set. Secondly, using the pre-trained network on large-scale datasets, such as VGG16, the bottleneck features in the small dataset are extracted and to be stored in two NumPy files as new training datasets and test datasets. Finally, training a fully connected network with the new datasets. In this paper, we use Kaggle famous Dogs vs Cats dataset as the experimental dataset, which is a two-category classification dataset.

• 전자통신동향분석
• /
• 제38권6호
• /
• pp.84-94
• /
• 2023

Open source has many advantages and is widely used in various fields. However, legal disputes regarding copyright and licensing of datasets and learning models have recently arisen in artificial intelligence developments. We examine how datasets affect artificial intelligence learning and services from the perspective of copyrighting and licensing when datasets are used for training models. The licensing conditions of datasets can lead to copyright infringement and license violation, thus determining the scope of disclosure and commercialization of the trained model. In addition, we examine related legal issues.

• 전기전자학회논문지
• /
• 제22권1호
• /
• pp.14-20
• /
• 2018

The purpose of training a convolutional neural network (CNN) is to obtain weight factors that give high classification accuracies. The initial values of hyper-parameters affect the training results, and it is important to train a CNN with a suitable hyper-parameter set of a learning rate, a batch size, the initialization of weight factors, and an optimizer. We investigate the effects of a single hyper-parameter while others are fixed in order to obtain a hyper-parameter set that gives higher classification accuracies and requires shorter training time using a proposed VGG-like CNN for training since the VGG is widely used. The CNN is trained for four datasets of CIFAR10, CIFAR100, GTSRB and DSDL-DB. The effects of the normalization and the data transformation for datasets are also investigated, and a training scheme using merged datasets is proposed.

• ETRI Journal
• /
• 제39권5호
• /
• pp.621-631
• /
• 2017

Abnormal samples are usually difficult to obtain in production systems, resulting in imbalanced training sample sets. Namely, the number of positive samples is far less than the number of negative samples. Traditional Support Vector Machine (SVM)-based anomaly detection algorithms perform poorly for highly imbalanced datasets: the learned classification hyperplane skews toward the positive samples, resulting in a high false-negative rate. This article proposes a new imbalanced SVM (termed ImSVM)-based anomaly detection algorithm, which assigns a different weight for each positive support vector in the decision function. ImSVM adjusts the learned classification hyperplane to make the decision function achieve a maximum GMean measure value on the dataset. The above problem is converted into an unconstrained optimization problem to search the optimal weight vector. Experiments are carried out on both Cloud datasets and Knowledge Discovery and Data Mining datasets to evaluate ImSVM. Highly imbalanced training sample sets are constructed. The experimental results show that ImSVM outperforms over-sampling techniques and several existing imbalanced SVM-based techniques.

• Journal of Multimedia Information System
• /
• 제9권2호
• /
• pp.75-86
• /
• 2022

The existing image quality assessment (IQA) datasets have a small number of samples. Some methods based on transfer learning or data augmentation cannot make good use of image quality-related features. A No Reference (NR)-IQA method based on multi-task training and quality awareness is proposed. First, single or multiple distortion types and levels are imposed on the original image, and different strategies are used to augment different types of distortion datasets. With the idea of weak supervision, we use the Full Reference (FR)-IQA methods to obtain the pseudo-score label of the generated image. Then, we combine the classification information of the distortion type, level, and the information of the image quality score. The ResNet50 network is trained in the pre-train stage on the augmented dataset to obtain more quality-aware pre-training weights. Finally, the fine-tuning stage training is performed on the target IQA dataset using the quality-aware weights to predicate the final prediction score. Various experiments designed on the synthetic distortions and authentic distortions datasets (LIVE, CSIQ, TID2013, LIVEC, KonIQ-10K) prove that the proposed method can utilize the image quality-related features better than the method using only single-task training. The extracted quality-aware features improve the accuracy of the model.

• Smart Structures and Systems
• /
• 제33권1호
• /
• pp.65-91
• /
• 2024

The use of five optimization techniques for the prediction of a strength-based concrete mixture's best-fit model is examined in this work. Five optimization techniques are utilized for this purpose: Slime Mold Algorithm (SMA), Black Hole Algorithm (BHA), Multi-Verse Optimizer (MVO), Vortex Search (VS), and Whale Optimization Algorithm (WOA). MATLAB employs a hybrid learning strategy to train an artificial neural network that combines least square estimation with backpropagation. Thus, 72 samples are utilized as training datasets and 31 as testing datasets, totaling 103. The multi-layer perceptron (MLP) is used to analyze all data, and results are verified by comparison. For training datasets in the best-fit models of SMA-MLP, BHA-MLP, MVO-MLP, VS-MLP, and WOA-MLP, the statistical indices of coefficient of determination (R²) in training phase are 0.9603, 0.9679, 0.9827, 0.9841 and 0.9770, and in testing phase are 0.9567, 0.9552, 0.9594, 0.9888 and 0.9695 respectively. In addition, the best-fit structures for training for SMA, BHA, MVO, VS, and WOA (all combined with multilayer perceptron, MLP) are achieved when the term population size was modified to 450, 500, 250, 150, and 500, respectively. Among all the suggested options, VS could offer a stronger prediction network for training MLP.

• 대한원격탐사학회지
• /
• 제38권4호
• /
• pp.395-409
• /
• 2022

The purpose of this study is to evaluate the classification performance and applicability when land cover datasets constructed for AI training are cross validation to other areas. For study areas, Gyeongsang-do and Jeolla-do in South Korea were selected as cross validation areas, and training datasets were obtained from AI-Hub. The obtained datasets were applied to the U-Net algorithm, a semantic segmentation algorithm, for each region, and the accuracy was evaluated by applying them to the same and other test areas. There was a difference of about 13-15% in overall classification accuracy between the same and other areas. For rice field, fields and buildings, higher accuracy was shown in the Jeolla-do test areas. For roads, higher accuracy was shown in the Gyeongsang-do test areas. In terms of the difference in accuracy by weight, the result of applying the weights of Gyeongsang-do showed high accuracy for forests, while that of applying the weights of Jeolla-do showed high accuracy for dry fields. The result of land cover classification, it was found that there is a difference in classification performance of existing datasets depending on area. When constructing land cover map for AI training, it is expected that higher quality datasets can be constructed by reflecting the characteristics of various areas. This study is highly scalable from two perspectives. First, it is to apply satellite images to AI study and to the field of land cover. Second, it is expanded based on satellite images and it is possible to use a large scale area and difficult to access.

• 대한원격탐사학회지
• /
• 제39권5_3호
• /
• pp.967-978
• /
• 2023

본 연구에서는 다양한 드론 영상 데이터셋을 효과적으로 학습하여 의미론적 분할의 정확도를 향상시키기 위한 combined segmentation network (CSN)를 제안하고 검증하였다. CSN은 세 가지 드론 데이터셋의 다양성을 고려하기 위하여 인코딩 영역의 전체를 공유하며, 디코딩 영역은 독립적으로 학습된다. CSN의 경우, 학습 시 모든 데이터셋에 대한 손실값을 고려하기 때문에 U-Net 및 pyramid scene parsing network (PSPNet)으로 단일 데이터셋을 학습할 때보다 학습 효율이 떨어졌다. 그러나 국내 자율주행 드론 영상에 CSN을 적용한 결과, CSN이 PSPNet에 비해 초기 학습 없이도 영상 내 화소를 적절한 클래스로 분류할 수 있는 것을 확인하였다. 본 연구를 통하여 CSN이 다양한 드론 영상 데이터셋을 효과적으로 학습하고 새로운 지역에 대한 객체 인식 정확성을 향상시키는 데 중요한 도구로써 활용될 수 있을 것으로 기대할 수 있다.

• ETRI Journal
• /
• 제41권2호
• /
• pp.242-253
• /
• 2019

We devise a layer-wise hint training method to improve the existing hint-based knowledge distillation (KD) training approach, which is employed for knowledge transfer in a teacher-student framework using a residual network (ResNet). To achieve this objective, the proposed method first iteratively trains the student ResNet and incrementally employs hint-based information extracted from the pretrained teacher ResNet containing several hint and guided layers. Next, typical softening factor-based KD training is performed using the previously estimated hint-based information. We compare the recognition accuracy of the proposed approach with that of KD training without hints, hint-based KD training, and ResNet-based layer-wise pretraining using reliable datasets, including CIFAR-10, CIFAR-100, and MNIST. When using the selected multiple hint-based information items and their layer-wise transfer in the proposed method, the trained student ResNet more accurately reflects the pretrained teacher ResNet's rich information than the baseline training methods, for all the benchmark datasets we consider in this study.