• International Journal of Computer Science & Network Security
• /
• 제24권2호
• /
• pp.89-94
• /
• 2024

Indian subcontinent is a birthplace of multilingual people where documents such as job application form, passport, number plate identification, and so forth is composed of text contents written in different languages/scripts. These scripts may be in the form of different indic numerals in a single document page. Due to this reason, building a generic recognizer that is capable of recognizing handwritten indic digits written by diverse writers is needed. Also, a lot of work has been done for various non-Indic numerals particularly, in case of Roman, but, in case of Indic digits, the research is limited. Moreover, most of the research focuses with only on MNIST datasets or with only single datasets, either because of time restraints or because the model is tailored to a specific task. In this work, a hybrid model is proposed to recognize all available indic handwritten digit images using the existing benchmark datasets. The proposed method bridges the automatically learnt features of Capsule Network with hand crafted Bag of Feature (BoF) extraction method. Along the way, we analyze (1) the successes (2) explore whether this method will perform well on more difficult conditions i.e. noise, color, affine transformations, intra-class variation, natural scenes. Experimental results show that the hybrid method gives better accuracy in comparison with Capsule Network.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권4호
• /
• pp.1275-1292
• /
• 2021

As a fine-grained classification problem, aspect-level sentiment classification predicts the sentiment polarity for different aspects in context. To address this issue, researchers have widely used attention mechanisms to abstract the relationship between context and aspects. Still, it is difficult to effectively obtain a more profound semantic representation, and the strong correlation between local context features and the aspect-based sentiment is rarely considered. In this paper, a hybrid attention capsule network for aspect-level sentiment classification (ABASCap) was proposed. In this model, the multi-head self-attention was improved, and a context mask mechanism based on adjustable context window was proposed, so as to effectively obtain the internal association between aspects and context. Moreover, the dynamic routing algorithm and activation function in capsule network were optimized to meet the task requirements. Finally, sufficient experiments were conducted on three benchmark datasets in different domains. Compared with other baseline models, ABASCap achieved better classification results, and outperformed the state-of-the-art methods in this task after incorporating pre-training BERT.

• International Journal of Computer Science & Network Security
• /
• 제21권11호
• /
• pp.338-344
• /
• 2021

The recently continuous enhancement and development in the biomedical side for the betterment of human life. The Wireless Body Area Networks is a significant tool for the current researcher to design and transfer data with greater data rates among the sensors and sensor nodes for biomedical applications. The core area for research in WBANs is power efficiency, battery-driven devices for health and medical, the Charging limitation is a major and serious problem for the WBANs.this research work is proposed to find out the optimal solution for battery-friendly technology. In this research we have addressed the solution to increasing the battery lifetime with variable data transmission rates from medical equipment as Wireless Endoscopy Capsules, this device will analyze a patient's inner body gastrointestinal tract by capturing images and visualization at the workstation. The second major issue is that the Wireless Endoscopy Capsule based systems are currently not used for clinical applications due to their low data rate as well as low resolution and limited battery lifetime, in case of these devices are more enhanced in these cases it will be the best solution for the medical applications. The main objective of this research is to power optimization by reducing the power consumption of the battery in the Wireless Endoscopy Capsule to make it battery-friendly. To overcome the problem we have proposed the algorithm for "Battery Friendly Algorithm" and we have compared the different frame rates of buffer sizes for Transmissions. The proposed Battery Friendly Algorithm is to send the images on average frame rate instead of transmitting the images on maximum or minimum frame rates. The proposed algorithm extends the battery lifetime in comparison with the previous baseline proposed algorithm as well as increased the battery lifetime of the capsule.

• Journal of Information Processing Systems
• /
• 제16권1호
• /
• pp.171-183
• /
• 2020

Capsule endoscopy is one of the increasingly demanded diagnostic methods among patients in recent years because of its ability to observe small intestine difficulties. It is often conducted for 12 to 14 hours, but significant frames constitute only 10% of whole frames. Thus, it has been designed to automatically acquire significant frames through deep learning. For example, studies to track the position of the capsule (stomach, small intestine, etc.) or to extract lesion-related information (polyps, etc.) have been conducted. However, although grouping or labeling the training images according to similar features can improve the performance of a learning model, various attributes (such as degree of wrinkles, presence of valves, etc.) are not considered in conventional approaches. Therefore, we propose a class-labeling method that can be used to design a learning model by constructing a knowledge model focused on main lesions defined in standard terminologies for capsule endoscopy (minimal standard terminology, capsule endoscopy structured terminology). This method enables the designing of a systematic learning model by labeling detailed classes through differentiation of similar characteristics.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2022년도 가을 학술논문 발표대회
• /
• pp.243-244
• /
• 2022

In this paper, experiments were conducted on signal amplification of polymer capsules for application to Ground Penetrating Radar so as to enable real-time monitoring of polymer capsules inside concrete using the Morphology Dependent Resonance phenomenon. A TEM CELL and a vector network analyzer were used to analyze the difference in resonance frequency depending on the material of the sphere and the presence or absence of fracture. In order to manufacture a capsule of a size that can be measured using millimeter waves used in GPR, we manufactured a capsule with a 3D printer and analyzed the effects of the presence or absence of coating and the size of the capsule on the resonance frequency. Resonant frequency or signal amplification is more affected by diameter than coating. The capsule showing the highest amplification is the resin-coated 50 mm diameter capsule with a 316-fold increase and the lowest capsule is the uncoated 10 mm diameter capsule with a signal amplification of 11.9 times. These results demonstrate the potential of GPR to measure the position and state of self-healing capsules, which are small-sized polymers, in real time using millimeter waves.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2022년도 추계학술대회
• /
• pp.2-5
• /
• 2022

사람의 체내 장기는 복잡한 구조로 되어있으며 특히, 소장은 길이가 약 7m 길이를 가지고 있어 내시경 검사가 쉽지 않고 내시경 검사 시 위험도가 높다. 현재는 캡슐 내시경으로 검사를 수행하고 있으며, 검사 시간이 매우 긴 편이다. 의사는 제거된 저장장치를 컴퓨터에 연결해 환자의 캡슐 내시경 영상을 저장 후 프로그램을 사용하여 판독하지만, 캡슐 내시경 검사 결과 영상 길이가 길어 판독 시간이 많이 소요된다. 또한 소장의 경우 융모에 의해 많은 굴곡이 존재해 검사 과정에서 영상의 폐색 영역이나 명암이 뚜렷이 나타나게 되어 검사 시 병변 및 이상징후에 관해 놓치는 경우가 발생할 수 있다. 본 논문에서는 의사의 영상 판독 시간 단축과 진단 신뢰도 향상을 위해 인공신경망을 이용한 소장 캡슐 내시경 병변 검사 보조 방법을 제공한다.

• ETRI Journal
• /
• 제37권3호
• /
• pp.637-645
• /
• 2015

Common-mode noise (CMN) is an unresolved problem in wireless capsule endoscopy (WCE) systems. In a WCE system, CMN originates from various electric currents found within the human body or external interference sources and causes critical demodulation performance degradation. The differential operation, a typical method for the removal of CMN rejection, can remove CMN by subtracting two signals simultaneously received by two reception sensors attached to a human body. However, when there is impedance mismatching between the two reception sensors, the differential operation method cannot completely remove CMN. Therefore, to overcome this problem, we propose an enhanced CMN rejection method. The proposed method performs not only subtraction but also addition between two received signals. Then a CMN ratio can be estimated by sufficient accumulation of division operation outcomes between the subtraction and addition outputs during the guard period. Finally, we can reject the residual CMN by combining the subtraction and addition outputs.

• 한국정밀공학회지
• /
• 제26권12호
• /
• pp.117-122
• /
• 2009

The hexagonal network-type PDMS microstructures were fabricated and they were employed to low-friction drag surfaces. While the lowest contact angle measured from the smooth surface was $108^{\circ}$ the highest contact angle measured from the microstructured surfaces was $145^{\circ}$ The moving speed of bullet-type capsule attached with a PDMS pad of smooth surface ($CA=108^{\circ}$) was 0.1261 m/s and that with a PDMS pad of microstructured surface ($CA=145^{\circ}$) was 0.1464 m/s. Compared with the smooth surface, the microstructured surface showed 16.1% higher moving speed. The network-type microstructures have a composite surface that is composed with air and PDMS solid. Therefore, the surface does not wet: rather water is lifted by the microstructures. Because of the composite surface, water shows slip-flow on the microstructures, and thus friction drag can be reduced.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제9권3호
• /
• pp.101-108
• /
• 2020

최근의 영상 처리 분야는 딥러닝 기법들의 성능이 입증됨에 따라 다양한 분야에서 이와 같은 기법들을 활용해 영상에 대한 분류, 분석, 검출 등을 수행하려는 시도가 활발하다. 그중에서도 의료 진단 보조 역할을 할 수 있는 의료 영상 분석 소프트웨어에 대한 기대가 증가하고 있는데, 본 연구에서는 데이터 셋이 방대하고 판단에 시간이 오래 걸리는 캡슐내시경 영상에 주목하였다. 본 논문의 목적은 캡슐내시경 영상의 판독에서 모든 환자에 대해 공통으로 수행되고, 판독하는 데 많은 시간을 차지하는 위장관 랜드마크를 구별하고 위장관 교차점을 추정하는 것이다. 이를 위해, 위장관 랜드마크를 식별할 수 있는 CNN 학습 모델을 설계하였으며, 이를 이용하여 결괏값을 필터링해 위장관 교차점을 추정하였다. 무작위로 환자 데이터를 샘플링한 모델을 이용해서 나온 결과를 필터링 후에 위장관 교차점을 추정하였을 때, 88% 환자는 위장에서 소장으로 변화하는 위장관 교차점(유문판) 의심 구역 안에 들어왔으며, 소장에서 대장으로 변화하는 위장관 교차점(회맹판)의 경우 100% 환자가 위장관 교차점 의심 구역 안에 들어온 것을 확인할 수 있었다. 100프레임 범위로 위장관 교차점 의심 구역을 찾을 수 있었으며, 판독자가 초당 10프레임의 속도로 판독을 진행한다면 10초안에 위장관 교차점을 찾아낼 수 있다.

• 한국가시화정보학회:학술대회논문집
• /
• 한국가시화정보학회 2005년도 추계학술대회 논문집
• /
• pp.89-92
• /
• 2005

A microcapsule for drug delivery was successfully produced by utilizing the flow-through droplet-based supramolecular self-assembly in a crossed microchannel network. The PS-b-PMMA block copolymer synthesized atom transfer radical polymerization (ATRP) was initially formed as microdroplets and after the evaporation process it turned to spherical capsule by polymer self-assembly of the micro domains. The characteristics were studied using various analysis methods.