• Journal of the Korea Society of Computer and Information
• /
• v.21 no.9
• /
• pp.11-18
• /
• 2016

In this paper, we propose a gas classification method using combined features for an electronic nose system that performs well even when some loss occurs in measuring data samples. We first divide the entire measurement for a data sample into three local sections, which are the stabilization, exposure, and purge; local features are then extracted from each section. Based on the discrimination analysis, measurements of the discriminative information amounts are taken. Subsequently, the local features that have a large amount of discriminative information are chosen to compose the combined features together with the global features that extracted from the entire measurement section of the data sample. The experimental results show that the combined features by the proposed method gives better classification performance for a variety of volatile organic compound data than the other feature types, especially when there is data loss.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.15 no.3
• /
• pp.298-305
• /
• 2012

In this paper, we propose an integrated algorithm for scene recognition, which has been a challenging computer vision problem, with application to mobile robot localization. The proposed scene recognition method utilizes SIFT and visual words as local-level features and GIST as a global-level feature. As local-level and global-level features complement each other, it results in improved performance for scene recognition. This improved algorithm is of low computational complexity and robust to image distortions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.9
• /
• pp.4412-4428
• /
• 2018

The purpose of this work is to solve the problem of representing an entire video using Convolutional Neural Network (CNN) features for human action recognition. Recently, due to insufficient GPU memory, it has been difficult to take the whole video as the input of the CNN for end-to-end learning. A typical method is to use sampled video frames as inputs and corresponding labels as supervision. One major issue of this popular approach is that the local samples may not contain the information indicated by the global labels and sufficient motion information. To address this issue, we propose a binary hashing method to enhance the local feature extractors. First, we extract the local features and aggregate them into global features using maximum/minimum pooling. Second, we use the binary hashing method to capture the motion features. Finally, we concatenate the hashing features with global features using different normalization methods to train the classifier. Experimental results on the JHMDB and MPII-Cooking datasets show that, for these new local features, binary hashing mapping on the sparsely sampled features led to significant performance improvements.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.4
• /
• pp.819-833
• /
• 2013

Computing global features based on local features using a bio-inspired framework has shown promising performance. However, for some tough applications with large intra-class variances, a single local feature is inadequate to represent all the attributes of the images. To integrate the complementary abilities of multiple local features, in this paper we have extended the efficacy of the bio-inspired framework, HMAX, to adapt heterogeneous features for global feature extraction. Given multiple global features, we propose an approach, designated as parameterized metric learning, for high dimensional feature fusion. The fusion parameters are solved by maximizing the canonical correlation with respect to the parameters. Experimental results show that our method achieves significant improvements over the benchmark bio-inspired framework, HMAX, and other related methods on the Caltech dataset, under varying numbers of training samples and feature elements.

• ETRI Journal
• /
• v.45 no.3
• /
• pp.505-517
• /
• 2023

Color, texture, and shape act as important information for images in human recognition. For content-based image retrieval, many studies have combined color, texture, and shape features to improve the retrieval performance. However, there have not been many powerful methods for combining all color, texture, and shape features. This study proposes a content-based image retrieval method that uses the combined local and global features of color, texture, and shape. The color features are extracted from the color autocorrelogram; the texture features are extracted from the magnitude of a complete local binary pattern and the Gabor local correlation revealing local image characteristics; and the shape features are extracted from singular value decomposition that reflects global image characteristics. In this work, an experiment is performed to compare the proposed method with those that use our partial features and some existing techniques. The results show an average precision that is 19.60% higher than those of existing methods and 9.09% higher than those of recent ones. In conclusion, our proposed method is superior over other methods in terms of retrieval performance.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.3
• /
• pp.894-907
• /
• 2022

The features extracted by convolutional neural networks are more descriptive of images than traditional features, and their convolutional layers are more suitable for retrieving images than are fully connected layers. The convolutional layer features will consume considerable time and memory if used directly to match an image. Therefore, this paper proposes a feature weighting and region integration method for convolutional layer features to form global feature vectors and subsequently use them for image matching. First, the 3D feature of the last convolutional layer is extracted, and the convolutional feature is subsequently weighted again to highlight the edge information and position information of the image. Next, we integrate several regional eigenvectors that are processed by sliding windows into a global eigenvector. Finally, the initial ranking of the retrieval is obtained by measuring the similarity of the query image and the test image using the cosine distance, and the final mean Average Precision (mAP) is obtained by using the extended query method for rearrangement. We conduct experiments using the Oxford5k and Paris6k datasets and their extended datasets, Paris106k and Oxford105k. These experimental results indicate that the global feature extracted by the new method can better describe an image.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.1
• /
• pp.105-125
• /
• 2024

Gliomas are the most common malignant brain tumor and cause the most deaths. Manual brain tumor segmentation is expensive, time-consuming, error-prone, and dependent on the radiologist's expertise and experience. Manual brain tumor segmentation outcomes by different radiologists for the same patient may differ. Thus, more robust, and dependable methods are needed. Medical imaging researchers produced numerous semi-automatic and fully automatic brain tumor segmentation algorithms using ML pipelines and accurate (handcrafted feature-based, etc.) or data-driven strategies. Current methods use CNN or handmade features such symmetry analysis, alignment-based features analysis, or textural qualities. CNN approaches provide unsupervised features, while manual features model domain knowledge. Cascaded algorithms may outperform feature-based or data-driven like CNN methods. A revolutionary cascaded strategy is presented that intelligently supplies CNN with past information from handmade feature-based ML algorithms. Each patient receives manual ground truth and four MRI modalities (T1, T1c, T2, and FLAIR). Handcrafted characteristics and deep learning are used to segment brain tumors in a Global Convolutional Neural Network (GCNN). The proposed GCNN architecture with two parallel CNNs, CSPathways CNN (CSPCNN) and MRI Pathways CNN (MRIPCNN), segmented BraTS brain tumors with high accuracy. The proposed model achieved a Dice score of 87% higher than the state of the art. This research could improve brain tumor segmentation, helping clinicians diagnose and treat patients.

• Journal of Institute of Control, Robotics and Systems
• /
• v.15 no.7
• /
• pp.734-741
• /
• 2009

A new localization algorithm is proposed for a fast moving mobile robot, which utilizes only one beacon and the global features of the differential-driving mobile robot. It takes a relatively long time to localize a mobile robot with active beacon sensors since the distance to the beacon is measured by the traveling time of the ultrasonic signal. When the mobile robot is moving slowly the measurement time does not yield a high error. At a higher mobile robot speed, however, the localization error becomes too large to locate the mobile robot. Therefore, in high-speed mobile robot operations, instead of using two or more active beacons for localization, only one active beacon and the global features of the mobile robot are used to localize the mobile robot in this research. The two global features are the radius and center of the rotational motion for the differential-driving mobile robot which generally describe motion of the mobile robot and are used for the trace prediction of the mobile robot. In high speed operations the localizer finds an intersection point of this predicted trace and a circle which is centered at the beacon and has the radius of the distance between the mobile robot and the beacon. This new approach resolves the large localization error caused by the high speed of the mobile robot. The performance of the new localization algorithm has been verified through the experiments with a high-speed mobile robot.

• International journal of advanced smart convergence
• /
• v.12 no.4
• /
• pp.434-442
• /
• 2023

The relationship between acute kidney injury (AKI) prediction and nephrotoxic drugs, or drugs that adversely affect kidney function, is one that has yet to be explored in the critical care setting. One contributing factor to this gap in research is the limited investigation of drug modalities in the intensive care unit (ICU) context, due to the challenges of processing prescription data into the corresponding drug representations and a lack in the comprehensive understanding of these drug representations. This study addresses this gap by proposing a novel approach that leverages patient prescription data as a modality to improve existing models for AKI prediction. We base our research on Electronic Health Record (EHR) data, extracting the relevant patient prescription information and converting it into the selected drug representation for our research, the extended-connectivity fingerprint (ECFP). Furthermore, we adopt a unique multimodal approach, developing machine learning models and 1D Convolutional Neural Networks (CNN) applied to clinical drug representations, establishing a procedure which has not been used by any previous studies predicting AKI. The findings showcase a notable improvement in AKI prediction through the integration of drug embeddings and other patient cohort features. By using drug features represented as ECFP molecular fingerprints along with common cohort features such as demographics and lab test values, we achieved a considerable improvement in model performance for the AKI prediction task over the baseline model which does not include the drug representations as features, indicating that our distinct approach enhances existing baseline techniques and highlights the relevance of drug data in predicting AKI in the ICU setting.

• Journal of Korea Multimedia Society
• /
• v.13 no.6
• /
• pp.875-883
• /
• 2010

In this paper, we present a secure biometric hashing scheme for face recognition by random fusion of global and local features. The Fourier-Mellin transform and Radon transform are adopted respectively to form specialized representation of global and local features, due to their invariance to geometric operations. The final biometric hash is securely generated by random weighting sum of both feature sets. A fourfold key is involved in our algorithm to ensure the security and privacy of biometric templates. The proposed biometric hash can be revocable and replaced by using a new key. Moreover, the attacker cannot obtain any information about the original biometric template without knowing the secret key. The experimental results confirm that our scheme has a satisfactory accuracy performance in terms of EER.