• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4426-4442
• /
• 2020

Since scene segmentation is becoming a hot topic in the field of autonomous driving and medical image analysis, researchers are actively trying new methods to improve segmentation accuracy. At present, the main issues in image semantic segmentation are intra-class inconsistency and inter-class indistinction. From our analysis, the lack of global information as well as macroscopic discrimination on the object are the two main reasons. In this paper, we propose a Densely connected residual Attention network (DA-Res2Net) which consists of a dense residual network and channel attention guidance module to deal with these problems and improve the accuracy of image segmentation. Specifically, in order to make the extracted features equipped with stronger multi-scale characteristics, a densely connected residual network is proposed as a feature extractor. Furthermore, to improve the representativeness of each channel feature, we design a Channel-Attention-Guide module to make the model focusing on the high-level semantic features and low-level location features simultaneously. Experimental results show that the method achieves significant performance on various datasets. Compared to other state-of-the-art methods, the proposed method reaches the mean IOU accuracy of 83.2% on PASCAL VOC 2012 and 79.7% on Cityscapes dataset, respectively.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.20 no.5
• /
• pp.133-138
• /
• 2010

We present a new multimodal biometric authentication method, which performs both feature-level fusion and decision-level fusion. After generating support vector machines for new features made by integrating face and voice features, the final decision for authentication is made by integrating decisions of face SVM classifier, voice SVM classifier and integrated features SVM clssifier. We justify our proposal by comparing our method with traditional one by experiments with XM2VTS multimodal database. The experiments show that our multilevel fusion algorithm gives higher recognition rate than the existing schemes.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.153-166
• /
• 2022

Identifying fine cracks in steel bridge facilities is a challenging task of structural health monitoring (SHM). This study proposed an end-to-end crack image segmentation framework based on a one-step Convolutional Neural Network (CNN) for pixel-level object recognition with high accuracy. To particularly address the challenges arising from small object detection in complex background, efforts were made in loss function selection aiming at sample imbalance and module modification in order to improve the generalization ability on complicated images. Specifically, loss functions were compared among alternatives including the Binary Cross Entropy (BCE), Focal, Tversky and Dice loss, with the last three specialized for biased sample distribution. Structural modifications with dilated convolution, Spatial Pyramid Pooling (SPP) and Feature Pyramid Network (FPN) were also performed to form a new backbone termed CrackDet. Models of various loss functions and feature extraction modules were trained on crack images and tested on full-scale images collected on steel box girders. The CNN model incorporated the classic U-Net as its backbone, and Dice loss as its loss function achieved the highest mean Intersection-over-Union (mIoU) of 0.7571 on full-scale pictures. In contrast, the best performance on cropped crack images was achieved by integrating CrackDet with Dice loss at a mIoU of 0.7670.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4456-4475
• /
• 2021

The system for re-identifying persons is used to find and verify the persons crossing through different spots using various cameras. Much research has been done to re-identify the person by utilising features with deep-learned or hand-crafted information. Deep learning techniques segregate and analyse the features of their layers in various forms, and the output is complex feature vectors. This paper proposes a distinctive framework called Integrated Level Feature Pattern (ILFP) framework, which integrates local and global features. A new deep learning architecture named modified XceptionNet (m-XceptionNet) is also proposed in this work, which extracts the global features effectively with lesser complexity. The proposed framework gives better performance in Rank1 metric for Market1501 (96.15%), CUHK03 (82.29%) and the newly created NEC01 (96.66%) datasets than the existing works. The mean Average Precision (mAP) calculated using the proposed framework gives 92%, 85% and 98%, respectively, for the same datasets.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.9
• /
• pp.2458-2482
• /
• 2023

Medical image segmentation techniques based on convolution neural networks indulge in feature extraction triggering redundancy of parameters and unsatisfactory target localization, which outcomes in less accurate segmentation results to assist doctors in diagnosis. In this paper, we propose a multi-level semantic-rich encoding-decoding network, which consists of a Pooling-Conv-Former (PCFormer) module and a Cbam-Dilated-Transformer (CDT) module. In the PCFormer module, it is used to tackle the issue of parameter explosion in the conservative transformer and to compensate for the feature loss in the down-sampling process. In the CDT module, the Cbam attention module is adopted to highlight the feature regions by blending the intersection of attention mechanisms implicitly, and the Dilated convolution-Concat (DCC) module is designed as a parallel concatenation of multiple atrous convolution blocks to display the expanded perceptual field explicitly. In addition, MultiHead Attention-DwConv-Transformer (MDTransformer) module is utilized to evidently distinguish the target region from the background region. Extensive experiments on medical image segmentation from Glas, SIIM-ACR, ISIC and LGG demonstrated that our proposed network outperforms existing advanced methods in terms of both objective evaluation and subjective visual performance.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.2
• /
• pp.182-189
• /
• 2002

In this paper, we research the content-based analysis and classification according to the composition of the feature parameters pool for the auditory signals to implement the auditory indexing and searching system. Auditory data is classified to the primitive various auditory types. we described the analysis and feature extraction method for the feature parameters available to the auditory data classification. And we compose the feature parameters pool in the indexing group unit, then compare and analysis the auditory data centering around the including level and indexing criterion into the audio categories. Based on this result, we composed the classification procedure and simulate the auditory data classification.

• Journal of Korea Multimedia Society
• /
• v.23 no.12
• /
• pp.1540-1551
• /
• 2020

Aiming at the problem that the existing human behavior recognition algorithm cannot fully utilize the multi-level spatio-temporal information of the network, a human behavior recognition algorithm based on a dense three-dimensional residual network is proposed. First, the proposed algorithm uses a dense block of three-dimensional residuals as the basic module of the network. The module extracts the hierarchical features of human behavior through densely connected convolutional layers; Secondly, the local feature aggregation adaptive method is used to learn the local dense features of human behavior; Then, the residual connection module is applied to promote the flow of feature information and reduced the difficulty of training; Finally, the multi-layer local feature extraction of the network is realized by cascading multiple three-dimensional residual dense blocks, and use the global feature aggregation adaptive method to learn the features of all network layers to realize human behavior recognition. A large number of experimental results on benchmark datasets KTH show that the recognition rate (top-l accuracy) of the proposed algorithm reaches 93.52%. Compared with the three-dimensional convolutional neural network (C3D) algorithm, it has improved by 3.93 percentage points. The proposed algorithm framework has good robustness and transfer learning ability, and can effectively handle a variety of video behavior recognition tasks.

• International Journal of Computer Science & Network Security
• /
• v.22 no.1
• /
• pp.167-174
• /
• 2022

In metaheuristic algorithms such as Genetic Algorithm (GA), initial population has a significant impact as it affects the time such algorithm takes to obtain an optimal solution to the given problem. In addition, it may influence the quality of the solution obtained. In the machine learning field, feature selection is an important process to attaining a good performance model; Genetic algorithm has been utilized for this purpose by scientists. However, the characteristics of Genetic algorithm, namely random initial population generation from a vector of feature elements, may influence solution and execution time. In this paper, the use of a statistical algorithm has been introduced (Chi2) for feature relevant checks where p-values of conditional independence were considered. Features with low p-values were discarded and subject relevant subset of features to Genetic Algorithm. This is to gain a level of certainty of the fitness of features randomly selected. An ensembled-based learning model for Hepatitis has been developed for Hepatitis C stage classification. 1385 samples were used using Egyptian-dataset obtained from UCI repository. The comparative evaluation confirms decreased in execution time and an increase in model performance accuracy from 56% to 63%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.12
• /
• pp.4568-4587
• /
• 2014

Authentication of videos and images based on the content is becoming an important problem in information security. Unfortunately, previous studies lack the consideration of Kerckhoffs's principle in order to achieve this (i.e., a cryptosystem should be secure even if everything about the system, except the key, is public knowledge). In this paper, a solution to the problem of finding a relationship between a frame's index and its content is proposed based on the creative utilization of a robust manifold feature. The proposed solution is based on a novel semi-fragile watermarking scheme for H.264/AVC video content authentication. At first, the input I-frame is partitioned for feature extraction and watermark embedding. This is followed by the temporal feature extraction using the Isometric Mapping algorithm. The frame index is included in the feature to produce the temporal watermark. In order to improve security, the spatial watermark will be encrypted together with the temporal watermark. Finally, the resultant watermark is embedded into the Discrete Cosine Transform coefficients in the diagonal positions. At the receiver side, after watermark extraction and decryption, temporal tampering is detected through a mismatch between the frame index extracted from the temporal watermark and the observed frame index. Next, the feature is regenerate through temporal feature regeneration, and compared with the extracted feature. It is judged through the comparison whether the extracted temporal watermark is similar to that of the original watermarked video. Additionally, for spatial authentication, the tampered areas are located via the comparison between extracted and regenerated spatial features. Experimental results show that the proposed method is sensitive to intentional malicious attacks and modifications, whereas it is robust to legitimate manipulations, such as certain level of lossy compression, channel noise, Gaussian filtering and brightness adjustment. Through a comparison between the extracted frame index and the current frame index, the temporal tempering is identified. With the proposed scheme, a solution to the Kerckhoffs's principle problem is specified.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.28 no.5
• /
• pp.39-47
• /
• 2000

This study is to investigate the conditions closely related to the establishment of vegetation in the riparian zone: the soil condition, an important factor along with climate and light. Especially, the soil structure of the microtopographical formations in the specific area known as the riparian microtopographical zone investigated. In addition, the effect of the riparian microtopographical features on the ground water level, soil moisture content, and vegetation was studied. The results of this study are as follows; 1) At all sample sites, below the sand layer, a gravel layer is always present. This is the result of past floods. 2) Although Salix koreensis experiences frequent disturbances such as increase in river level and floods, this vegetation establishes itself in the most secure are in the microtopographical zone. 3) The growth of Phragmites japonica is closely related to the underground water level. 4) It is clear that Miscanthus sacchariflorus grows concentrated in dry areas. 5) The soil accumulation conditions differ according to the soil moisture content of each microtopgraphical feature. Accordingly, the moisture content of the soil is clearly different within the microtopographical zone. The continuous and long-term investigation and research on the relation of riparian reproduction and the relevance with location surrounding factors are necessary in the future.