Optimal Viewpoint Extraction of 3D Model Based on AdaBoost Iterative Algorithm

Dong Tao 보안공학연구지원센터 2016 International Journal of Multimedia and Ubiquitous Vol.11 No.1

According to the limitations of a single measurement algorithm in the current 3D models’ viewpoint extraction, this essay puts forward a viewpoint extraction algorithm based on AdaBoost iterative algorithm, which can make the features adaptive automatically. It, firstly, extracts 3D models’ feature descriptor and feature vector in the model library and adopts AdaBoost iterative algorithm to establish rules about classification and matching from geometric features and various viewpoint extraction algorithm; then, it constructs decision classifier in order to extract optimal viewpoint. In query process, the model obtains viewpoint extraction algorithm which can suit its geometric feature through decision classifier and then gets its best view by calculation. The experimental result shows this algorithm extraction effect is superior to the one by a single measurement algorithm.

시간 간격 특징 벡터를 이용한 AdaBoost 기반 제스처 인식

황승준 ( Seung Jun Hwang ),안광표 ( Gwang Pyo Ahn ),박승제 ( Seung Je Park ),백중환 ( Joong Hwan Baek ) 한국항행학회 2013 韓國航行學會論文誌 Vol.17 No.2

본 논문에서는 키넥트 센서를 이용한 AdaBoost 기반 제스처 인식에 관한 알고리즘을 제안한다. 최근 스마트 TV에 대한 보급으로 관련 산업이 주목받고 있다. 기존 리모컨을 이용하여 TV를 컨트롤 하던 시대에서 벗어나 제스처를 이용하여 TV를 컨트롤 할 수 있는 새로운 접근을 제안한다. AdaBoost 학습 모델에 신체 정규화 된 시간 간격 특징 벡터의 집합을 특징 패턴으로 하여, 속도가 다른 동작들을 인식할 수 있도록 하였다. 또한 속도가 다른 다양한 제스처를 인식하기 위해 다중 AdaBoost 알고리즘을 적용하였다. 제안된 알고리즘을 실제 동영상 플레이어와 연결하여 적용하였고, 실험 후 좌표 변화를 이용한 알고리즘에 비해 정확도가 향상되었음을 확인하였다. The task of 3D gesture recognition for controlling equipments is highly challenging due to the propagation of 3D smart TV recently. In this paper, the AdaBoost algorithm is applied to 3D gesture recognition by using Kinect sensor. By tracking time interval trajectory of hand, wrist and arm by Kinect, AdaBoost algorithm is used to train and classify 3D gesture. Experimental results demonstrate that the proposed method can successfully extract trained gestures from continuous hand, wrist and arm motion in real time.

에이다부스트 알고리즘을 이용한 인체 영역의 강인한 검출

장석우(Seok-Woo Jang),변시우(Siwoo Byun) 한국산학기술학회 2016 한국산학기술학회논문지 Vol.17 No.11

최근 들어, 나체 사진이나 그림과 같은 유해한 영상 콘텐츠가 쉽게 유통 및 보급되고 있는 실정이어서 이런 유해영상 콘텐츠를 효과적으로 검출하고 필터링하기 위한 연구 방법들이 지속적으로 소개되고 있다. 따라서 본 논문에서는 입력되는 칼라 영상으로부터 영상의 유해성을 나타내는 요소인 사람의 배꼽 영역을 하르-라이크(Haar-like) 특징과 에이다부스트(Adaboost) 알고리즘을 이용하여 강인하게 검출하는 새로운 접근 방법을 제안한다. 본 논문에서 제안된 방법에서는 먼저 입력 영상으로부터 색상 정보를 이용하여 사람의 유두 영역을 검출하고, 검출된 유두 영역과의 위치 정보를 사용하여 배꼽의 후보 영역을 검출한다. 그런 다음, 하르-라이크 특징과 에이다부스트 알고리즘을 이용한 필터링을 통해 실제 배꼽 영역들만을 검출한다. 실험 결과에서는 제안된 방법이 입력되는 칼라 영상으로부터 배꼽 영역을 기존의 방법보다 1.6% 더 정확하게 추출한다는 것을 보여준다. 본 논문에서 제안된 배꼽 영역 검출 알고리즘은 2 차원이나 3 차원의 유해 콘텐츠 검출 및 필터링과 관련된 여러 가지 응용 분야에서 매우 효과적으로 이용될 것으로 기대된다. Recently, harmful content (such as images and photos of nudes) has been widely distributed. Therefore, there have been various studies to detect and filter out such harmful image content. In this paper, we propose a new method using Haar-like features and an AdaBoost algorithm for robustly extracting navel areas in a color image. The suggested algorithm first detects the human nipples through color information, and obtains candidate navel areas with positional information from the extracted nipple areas. The method then selects real navel regions based on filtering using Haar-like features and an AdaBoost algorithm. Experimental results show that the suggested algorithm detects navel areas in color images 1.6 percent more robustly than an existing method. We expect that the suggested navel detection algorithm will be usefully utilized in many application areas related to 2D or 3D harmful content detection and filtering.

모노 카메라 영상기반 시간 간격 윈도우를 이용한 광역 및 지역 특징 벡터 적용 AdaBoost기반 제스처 인식

황승준(Seung-Jun Hwang),고하윤(Ha-Yoon Ko),백중환(Joong-Hwan Baek) 한국정보통신학회 2018 한국정보통신학회논문지 Vol.22 No.3

최근 안드로이드, iOS 등의 셋톱박스 기반의 스마트 TV에 대한 보급에 따라 제스처로 TV를 컨트롤 할 수 있는 새로운 접근을 제안한다. 본 논문에서는 모노 카메라 센서를 이용한 AdaBoost 기반 제스처 인식에 관한 알고리즘을 제안한다. 우선, 신체 좌표 추출을 위해 가우시안 배경 제거 및 Camshift 기반 자세 추적 및 추정 알고리즘을 사용한다. AdaBoost 학습 모델을 신체 정규화된 광역 및 지역 특징 벡터의 집합을 특징 패턴으로 하여, 속도가 다른 동작들을 인식할 수 있도록 하였다. 또한 속도가 다른 다양한 제스처를 인식하기 위해 다중 AdaBoost 알고리즘을 적용하였다. CART 알고리즘을 이용하여 성공적인 중요 특징 벡터를 확인하고 중요도가 낮은 특징벡터를 제거하는 방식을 적용하면서 분류 성공률이 높은 최적의 특징 벡터를 탐색하였다. 그 결과 24개의 주성분 특징 벡터를 찾았으며, 기존 알고리즘에 비해 낮은 오분류율(3.73%)과 높은 인식률(95.17%)을 지닌 특징 벡터 및 분류기를 설계하였다. Recently, the spread of smart TV based Android iOS Set Top box has become common. This paper propose a new approach to control the TV using gestures away from the era of controlling the TV using remote control. In this paper, the AdaBoost algorithm is applied to gesture recognition by using a mono camera. First, we use Camshift-based Body tracking and estimation algorithm based on Gaussian background removal for body coordinate extraction. Using global and local feature vectors, we recognized gestures with speed change. By tracking the time interval trajectories of hand and wrist, the AdaBoost algorithm with CART algorithm is used to train and classify gestures. The principal component feature vector with high classification success rate is searched using CART algorithm. As a result, 24 optimal feature vectors were found, which showed lower error rate (3.73%) and higher accuracy rate (95.17%) than the existing algorithm.

AdaBoost 기반 알고리즘을 이용한 실시간 얼굴 검출 및 추적

김은경(Eun-Kyoung Kim),김현주(HyunJu Kim),조현미(Hyun Mi Jo),이용환(Yong-Hwan Lee) 한국정보기술학회 2011 Proceedings of KIIT Conference Vol.2011 No.5

본 논문은 얼굴 검출 및 추적을 하기 위해 적분이미지로부터 추출한 Haar-like Feature를 이용한 AdaBoost 알고리즘을 사용하였으며, 추적이 완료됨에 따라 해당 이미지를 저장하도록 설계하였다. AdaBoost 알고리즘의 원리는 학습초기에 샘플 데이터의 가중치는 동일한 상태에서 시작하지만 매 라운드가 진행될수록 이전의 분류규칙을 적용하여 얻은 관측값으로 오분류 데이터에는 높은 가중치를 부여하고 반대로 정분류된 데이터에는 낮은 가중치를 부여하는 방식으로 샘플 데이터의 분포를 재조정한다. AdaBoost 알고리즘 기반 얼굴검출 기법은 빠른 검출속도와 높은 검출 결과를 보여주며, 얼굴 검출은 얼굴과 비얼굴 영상으로 구성된 학습용 영상집합에서 추출한 특징집합을 이용하여 학습된다.

Performance Comparison of Sentiment Lexicons in Predicting American Citizen's Sentiment for the Apartment Rents

김윤기 한국지적학회 2020 한국지적학회지 Vol.36 No.1

The primary purpose of this study is to identify which sentiment lexicon performs better in predicting citizens' sentiment about apartment rents. To accomplish this research purpose, I performed sentiment analysis based on sentiment lexicon and sentiment analysis based on the machine learning algorithm simultaneously. As a result of the analysis, the AFINN Sentiment Lexicon turned out to be the best performing sentiment lexicon in all performance metrics categories. Also, the performance of each machine learning classifier showed a significant difference in classifying and predicting citizens' sentiment on apartment rents by sentiment lexicon. In other words, AdaBoost in the AFINN Sentiment Lexicon and the Opinion Lexicon and Random Forest in the NRC Emotion Lexicon proved to be the best algorithms. This study can contribute to the research of this area because it uses sentiment analysis based on the sentiment lexicon and sentiment analysis based on the machine learning algorithm simultaneously in classifying and predicting citizens' sentiment about apartment rents. Despite these advantages, this research has limitations in that it uses only a limited range of sentiment lexicons and machine learning algorithms. Therefore, future studies will have to use more sentiment lexicons and machine learning algorithms to classify and predict citizens' sentiment about apartment rents accurately.

에지 분석과 에이다부스트 알고리즘을 이용한 차량검출

송광열(Gwangyul Song),이기용(Kiyong Lee),이준웅(Joonwoong Lee),윤팔주(Paljoo Yoon) 한국자동차공학회 2007 한국자동차공학회 Symposium Vol.- No.-

This paper proposes an algorithm capable of detecting vehicles in front or in rear using a monocular camera installed in a vehicle. The vehicle detection has been regarded as an important part of intelligent vehicle technologies. The proposed algorithm is mainly composed of two parts: 1)hypothesis generation of vehicles, and 2)hypothesis verification. The hypotheses of vehicles are generated by the analysis of vertical and horizontal edges and the detection of symmetry axis. The hypothesis verification, which determines vehicles among hypotheses, is done by the AdaBoost algorithm. The proposed algorithm is proven to be effective through experiments performed on various images captured on the roads.

A Fast and Accurate Face Tracking Scheme by using Depth Information in Addition to Texture Information

Kim, Dong-Wook,Kim, Woo-Youl,Yoo, Jisang,Seo, Young-Ho The Korean Institute of Electrical Engineers 2014 Journal of Electrical Engineering & Technology Vol.9 No.2

This paper proposes a face tracking scheme that is a combination of a face detection algorithm and a face tracking algorithm. The proposed face detection algorithm basically uses the Adaboost algorithm, but the amount of search area is dramatically reduced, by using skin color and motion information in the depth map. Also, we propose a face tracking algorithm that uses a template matching method with depth information only. It also includes an early termination scheme, by a spiral search for template matching, which reduces the operation time with small loss in accuracy. It also incorporates an additional simple refinement process to make the loss in accuracy smaller. When the face tracking scheme fails to track the face, it automatically goes back to the face detection scheme, to find a new face to track. The two schemes are experimented with some home-made test sequences, and some in public. The experimental results are compared to show that they outperform the existing methods in accuracy and speed. Also we show some trade-offs between the tracking accuracy and the execution time for broader application.

A Fast and Accurate Face Tracking Scheme by using Depth Information in Addition to Texture Information

Dong-Wook Kim,Woo-Youl Kim,Jisang Yoo,Young-Ho Seo 대한전기학회 2014 Journal of Electrical Engineering & Technology Vol.9 No.2

This paper proposes a face tracking scheme that is a combination of a face detection algorithm and a face tracking algorithm. The proposed face detection algorithm basically uses the Adaboost algorithm, but the amount of search area is dramatically reduced, by using skin color and motion information in the depth map. Also, we propose a face tracking algorithm that uses a template matching method with depth information only. It also includes an early termination scheme, by a spiral search for template matching, which reduces the operation time with small loss in accuracy. It also incorporates an additional simple refinement process to make the loss in accuracy smaller. When the face tracking scheme fails to track the face, it automatically goes back to the face detection scheme, to find a new face to track. The two schemes are experimented with some home-made test sequences, and some in public. The experimental results are compared to show that they outperform the existing methods in accuracy and speed. Also we show some trade-offs between the tracking accuracy and the execution time for broader application.

Pedestrian Detection Approach Based on Modified Haar-Like Features and AdaBoost

Van-Dung Hoang,Andrey Vavilin,Kang-Hyun Jo 제어로봇시스템학회 2012 제어로봇시스템학회 국제학술대회 논문집 Vol.2012 No.10

Pedestrian detection is an important task in many applications such as intelligent transportation systems, image retrieval, surveillance systems, automated personal assistance, etc. This paper proposes a set of modified Haar-like features that have parallelogram shapes. Using the proposed feature descriptors to develop a rapid detection system for pedestrian detection based on decision tree structure used boosting algorithm. The experimental results showed that the proposed method could produce high accuracy detection rate with lower false positive rate and higher recall rate than original Haar-like features and it is efficiency with different resolutions and gestures under a variety of backgrounds as well as lighting.