http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
기민송(Minsong Ki),최영우(Yeongwoo Choi) 한국디지털콘텐츠학회 2020 한국디지털콘텐츠학회논문지 Vol.21 No.2
We need research to detect human extreme pains using vision technologies in various environments for safety. For example, a car driver may be in a state of emergencies such as sudden pain or cardiac arrest. Therefore, we propose an emergency detection system based on human facial expressions to detect human emergent states. We organize the data for painful facial expression classes separately and propose a modified LeNet to use as a baseline. We add a resampling process to solve the noise in the training data. In addition, for Painful class with few samples and difficult to classify, we apply ring loss with softmax for clustering by facial expression class in feature space. We show accuracies of 63.3% and 60.4% for validation and testset with 8 expression classes both from 7 expression classes of FER2013[1] and an added pain class extracted from Pain Expression [4]. These results can hopefully be used to develop a system that can prevent terrible car accidents due to a sudden pain of the car drivers.
후미등 하단 학습기반의 차종에 무관한 전방 차량 검출 시스템
기민송(Minsong Ki),곽수영(Sooyeong Kwak),변혜란(Hyeran Byun) 한국방송·미디어공학회 2016 방송공학회논문지 Vol.21 No.4
Recently, there are active studies on a forward collision warning system to prevent the accidents and improve convenience of drivers. For collision evasion, the vehicle detection system is required. In general, existing learning-based vehicle detection methods use the entire appearance of the vehicles from rear-view images, so that each vehicle types should be learned separately since they have distinct rear-view appearance regarding the types. To overcome such shortcoming, we learn Haar-like features from the lower part of the vehicles which contain tail lights to detect vehicles leveraging the fact that the lower part is consistent regardless of vehicle types. As a verification procedure, we detect tail lights to distinguish actual vehicles and non-vehicles. If candidates are too small to detect the tail lights, we use HOG(Histogram Of Gradient) feature and SVM(Support Vector Machine) classifier to reduce false alarms. The proposed forward vehicle detection method shows accuracy of 95% even in the complicated images with many buildings by the road, regardless of vehicle types.