• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권6호
• /
• pp.1516-1529
• /
• 2023

In this paper, a motor gearbox fault-detection system based on a hierarchical flow-based model is proposed. The proposed system is used for the anomaly detection of a motion sound-based actuator module. The proposed flow-based model, which is a generative model, learns by directly modeling a data distribution function. As the objective function is the maximum likelihood value of the input data, the training is stable and simple to use for anomaly detection. The operation sound of a car's side-view mirror motor is converted into a Mel-spectrogram image, consisting of a folding signal and an unfolding signal, and used as training data in this experiment. The proposed system is composed of an encoder and a decoder. The data extracted from the layer of the pretrained feature extractor are used as the decoder input data in the encoder. This information is used in the decoder by performing an interlayer cross-scale convolution operation. The experimental results indicate that the context information of various dimensions extracted from the interlayer hierarchical data improves the defect detection accuracy. This paper is notable because it uses acoustic data and a normalizing flow model to detect outliers based on the features of experimental data.

• 한국지능시스템학회논문지
• /
• 제25권5호
• /
• pp.437-443
• /
• 2015

이 논문에서는 가시광선 얼굴영상과 그로부터 예측한 열 적외선 텍스처의 데이터 융합에 의한 얼굴인식 방법에 관하여 연구하였다. 제안하는 얼굴인식 기법은 가시광선 얼굴영상과 열 적외선 텍스처를 PCA에 의하여 낮은 차원의 특징공간에서 특징벡터로 변환한 다음, 다층 신경회로망을 사용하여 가시광선 영상 특징으로부터 얼굴의 열적외선 특징을 예측하여 열 적외선 텍스처를 생성하였다. 학습과정에서는 주어진 개체로부터 획득한 한 쌍의 가시광선 및 열 적외선 영상에 대해서 PCA를 이용하여 낮은 차원의 특징공간으로 변환한 다음, 가시광선 영상특징으로부터 열 분포 특징으로 매핑시키는 비선형 함수에 해당하는 신경회로망의 내부 파라미터를 결정한다. 학습된 신경회로망은 입력 가시광선 얼굴 특징으로부터 열 에너지 분포 특성의 PCA계수를 예측하고, 이로부터 열 적외선 텍스처를 생성한다. 대표적인 두 가지 얼굴인식 알고리즘 Eigenfaces와 Fisherfaces을 사용하여 NIST/Equinox 데이터베이스에 대하여 얼굴인식에 관한 실험을 수행하였다. 예측한 열 적외선 텍스처와 가시광선 얼굴영상의 데이터 융합결과는 가시광선 얼굴영상만을 사용한 경우에 비해서 얼굴인식의 성능이 개선되었음을 수신자 조작특성 (ROC) 및 첫 번째 매칭성능에 의하여 검증하였다.

• Journal of Multimedia Information System
• /
• 제8권3호
• /
• pp.175-182
• /
• 2021

Oral cancer is ranked second most diagnosed cancer among Indian population and ranked sixth all around the world. Oral cancer is one of the deadliest cancers with high mortality rate and very less 5-year survival rates even after treatment. It becomes necessary to detect oral malignancies as early as possible so that timely treatment may be given to patient and increase the survival chances. In recent years deep learning based frameworks have been proposed by many researchers that can detect malignancies from medical images. In this paper we have proposed a deep learning-based framework which detects oral cancer from histopathology images very efficiently. We have designed our model to split the color channels and extract deep features from these individual channels rather than single combined channel with the help of Efficient NET B3. These features from different channels are fused by using feature fusion module designed as a layer and placed before dense layers of Efficient NET. The experiments were performed on our own dataset collected from hospitals. We also performed experiments of BreakHis, and ICML datasets to evaluate our model. The results produced by our model are very good as compared to previously reported results.

• 한국군사과학기술학회지
• /
• 제14권4호
• /
• pp.539-547
• /
• 2011

In this paper, a method of generating path planning map is developed using digitalized geographic information such as FDB(Feature DataBase). FDB is widely used by the Army and needs to be applied to all weapon systems of newly developed. For the autonomous navigation of a robot, it is necessary to generate a path planning map by which a global path can be optimized. First, data included in FDB is analyzed in order to identify meaningful layers and attributes of which information can be used to generate the path planning map. Then for each of meaningful layers identified, a set of values of attributes in the layer is converted into the traverse cost using a matching table in which any combination of attribute values are matched into the corresponding traverse cost. For a certain region that is gridded, i.e., represented by a grid map, the traverse cost is extracted in a automatic manner for each gird of the region to generate the path planning map. Since multiple layers may be included in a single grid, an algorithm is developed to fusion several traverse costs. The proposed method is tested using a experimental program. Test results show that it can be a viable tool for generating the path planning map in real-time. The method can be used to generate other kinds of path planning maps using the digitalized geographic information as well.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권12호
• /
• pp.3976-3990
• /
• 2022

Currently, cross-attention is widely used in Siamese trackers to replace traditional correlation operations for feature fusion between template and search region. The former can establish a similar relationship between the target and the search region better than the latter for robust visual object tracking. But existing trackers using cross-attention only focus on rich semantic information of high-level features, while ignoring the appearance information contained in low-level features, which makes trackers vulnerable to interference from similar objects. In this paper, we propose a Multi-level Cross-attention Siamese network(MCSiam) to aggregate the semantic information and appearance information at the same time. Specifically, a multi-level cross-attention module is designed to fuse the multi-layer features extracted from the backbone, which integrate different levels of the template and search region features, so that the rich appearance information and semantic information can be used to carry out the tracking task simultaneously. In addition, before cross-attention, a target-aware module is introduced to enhance the target feature and alleviate interference, which makes the multi-level cross-attention module more efficient to fuse the information of the target and the search region. We test the MCSiam on four tracking benchmarks and the result show that the proposed tracker achieves comparable performance to the state-of-the-art trackers.