• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제2권2호
• /
• pp.120-133
• /
• 2008

This paper presents a novel approach for facial motion tracking and facial expression cloning to create a realistic facial animation of a 3D avatar. The exact head pose estimation and facial expression tracking are critical issues that must be solved when developing vision-based computer animation. In this paper, we deal with these two problems. The proposed approach consists of two phases: dynamic head pose estimation and facial expression cloning. The dynamic head pose estimation can robustly estimate a 3D head pose from input video images. Given an initial reference template of a face image and the corresponding 3D head pose, the full head motion is recovered by projecting a cylindrical head model onto the face image. It is possible to recover the head pose regardless of light variations and self-occlusion by updating the template dynamically. In the phase of synthesizing the facial expression, the variations of the major facial feature points of the face images are tracked by using optical flow and the variations are retargeted to the 3D face model. At the same time, we exploit the RBF (Radial Basis Function) to deform the local area of the face model around the major feature points. Consequently, facial expression synthesis is done by directly tracking the variations of the major feature points and indirectly estimating the variations of the regional feature points. From the experiments, we can prove that the proposed vision-based facial expression cloning method automatically estimates the 3D head pose and produces realistic 3D facial expressions in real time.

• Journal of Astronomy and Space Sciences
• /
• 제35권4호
• /
• pp.263-277
• /
• 2018

This paper presents a vision-based relative pose estimation algorithm and its validation through both numerical and hardware experiments. The algorithm and the hardware system were simultaneously designed considering actual experimental conditions. Two estimation techniques were utilized to estimate relative pose; one was a nonlinear least square method for initial estimation, and the other was an extended Kalman Filter for subsequent on-line estimation. A measurement model of the vision sensor and equations of motion including nonlinear perturbations were utilized in the estimation process. Numerical simulations were performed and analyzed for both the autonomous docking and formation flying scenarios. A configuration of LED-based beacons was designed to avoid measurement singularity, and its structural information was implemented in the estimation algorithm. The proposed algorithm was verified again in the experimental environment by using the Autonomous Spacecraft Test Environment for Rendezvous In proXimity (ASTERIX) facility. Additionally, a laser distance meter was added to the estimation algorithm to improve the relative position estimation accuracy. Throughout this study, the performance required for autonomous docking could be presented by confirming the change in estimation accuracy with respect to the level of measurement error. In addition, hardware experiments confirmed the effectiveness of the suggested algorithm and its applicability to actual tasks in the real world.

• 한국멀티미디어학회논문지
• /
• 제24권5호
• /
• pp.659-666
• /
• 2021

With the continuous development of deep learning, human behavior recognition algorithms have achieved good results. However, in a multi-person recognition environment, the complex behavior environment poses a great challenge to the efficiency of recognition. To this end, this paper proposes a multi-person pose estimation model. First of all, the human detectors in the top-down framework mostly use the two-stage target detection model, which runs slow down. The single-stage YOLOv3 target detection model is used to effectively improve the running speed and the generalization of the model. Depth separable convolution, which further improves the speed of target detection and improves the model's ability to extract target proposed regions; Secondly, based on the feature pyramid network combined with context semantic information in the pose estimation model, the OHEM algorithm is used to solve difficult key point detection problems, and the accuracy of multi-person pose estimation is improved; Finally, the Euclidean distance is used to calculate the spatial distance between key points, to determine the similarity of postures in the frame, and to eliminate redundant postures.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2014년도 춘계학술발표대회
• /
• pp.904-907
• /
• 2014

In this work we present a robust and fast approach to estimate 3D vehicle pose that can provide results under a specific traffic surveillance conditions. Such limitations are expressed by single fixed CCTV camera that is located relatively high above the ground, its pitch axes is parallel to the reference plane and the camera focus assumed to be known. The benefit of our framework that it does not require prior training, camera calibration and does not heavily rely on 3D model shape as most common technics do. Also it deals with a bad shape condition of the objects as we focused on low resolution surveillance scenes. Pose estimation task is presented as PnP problem to solve it we use well known "POSIT" algorithm [1]. In order to use this algorithm at least 4 non coplanar point's correspondence is required. To find such we propose a set of techniques based on model and scene geometry. Our framework can be applied in real time video sequence. Results for estimated vehicle pose are shown in real image scene.

• 한국컴퓨터정보학회논문지
• /
• 제28권9호
• /
• pp.17-25
• /
• 2023

본 논문에서는 매개변수가 더 적고, 빠르게 추정 가능한 MobileViT 기반 모델을 통해 사람 자세 추정 과업을 수행할 수 있는 모델을 제안한다. 기반 모델은 합성곱 신경망의 특징과 Vision Transformer의 특징이 결합한 구조를 통해 경량화된 성능을 입증한다. 본 연구에서 주요 매커니즘이 되는 Transformer는 그 기반의 모델들이 컴퓨터 비전 분야에서도 합성곱 신경망 기반의 모델들 대비 더 나은 성능을 보이며, 영향력이 커지게 되었다. 이는 사람 자세 추정 과업에서도 동일한 상황이며, Vision Transformer기반의 ViTPose가 COCO, OCHuman, MPII 등 사람 자세 추정 벤치마크에서 모두 최고 성능을 지키고 있는 것이 그 적절한 예시이다. 하지만 Vision Transformer는 매개변수의 수가 많고 상대적으로 많은 연산량을 요구하는 무거운 모델 구조를 가지고 있기 때문에, 학습에 있어 사용자에게 많은 비용을 야기시킨다. 이에 기반 모델은 Vision Transformer가 많은 계산량을 요구하는 부족한 Inductive Bias 계산 문제를 합성곱 신경망 구조를 통한 Local Representation으로 극복하였다. 최종적으로, 제안 모델은 MS COCO 사람 자세 추정 벤치마크에서 제공하는 Validation Set으로 ViTPose 대비 각각 5분의 1과 9분의 1만큼의 3.28GFLOPs, 972만 매개변수를 나타내었고, 69.4 Mean Average Precision을 달성하여 상대적으로 우수한 성능을 보였다.

• 한국전자통신학회논문지
• /
• 제19권1호
• /
• pp.309-316
• /
• 2024

얼굴 인식 기술은 다양한 분야에서 활용되고 있지만, 이는 사진 스푸핑과 같은 위조 공격에 취약하다는 문제를 가지고 있다. 이를 극복하기 위한 여러 연구가 진행되고 있지만, 대부분은 멀티모달 카메라와 같은 특별한 장비를 장착하거나 고성능 환경에서 동작하는 것을 전제로 하고 있다. 본 연구는 얼굴 인식 위조 공격 문제를 해결하기 위해, 특별한 장비 없이 일반적인 웹캠에서 동작할 수 있는 LH-FAS v2를 제안한다. 제안된 방법에서는, 머리 자세 추정에는 FSA-Net을, 얼굴 식별에는 ArcFace를 활용하여 사진 스푸핑 여부를 판별한다. 실험을 위해, 사진 스푸핑 공격 비디오로 구성된 VD4PS 데이터셋을 제시하였으며, 이를 통해 LH-FAS v2의 균형 잡힌 정확도와 속도를 확인하였다. 본 방법은 향후 사진 스푸핑 방어에 효과적일 것으로 기대한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권8호
• /
• pp.3136-3150
• /
• 2015

A vision-based 3D tracking of articulated human hand is one of the major issues in the applications of human computer interactions and understanding the control of robot hand. This paper presents an improved approach for tracking and recovering the 3D position and orientation of a human hand using the Kinect sensor. The basic idea of the proposed method is to solve an optimization problem that minimizes the discrepancy in 3D shape between an actual hand observed by Kinect and a hypothesized 3D hand model. Since each of the 3D hand pose has 23 degrees of freedom, the hand articulation tracking needs computational excessive burden in minimizing the 3D shape discrepancy between an observed hand and a 3D hand model. For this, we first created a 3D hand model which represents the hand with 17 different parts. Secondly, Random Forest classifier was trained on the synthetic depth images generated by animating the developed 3D hand model, which was then used for Haar-like feature-based classification rather than performing per-pixel classification. Classification results were used for estimating the joint positions for the hand skeleton. Through the experiment, we were able to prove that the proposed method showed improvement rates in hand part recognition and a performance of 20-30 fps. The results confirmed its practical use in classifying hand area and successfully tracked and recovered the 3D hand pose in a real time fashion.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제11권10호
• /
• pp.419-426
• /
• 2022

본 논문에서는 다양한 운동 모션에서 3차원 사람 자세 추정 모델의 정확도를 향상하는 방법을 제안한다. 기존의 사람 자세 추정 모델은 사람의 자세를 추정할 때 좌표 오차를 유발하는 흔들림, 반전, 교환, 오검출 등의 문제가 발생한다. 이러한 문제는 사람 자세 추정 모델의 정확한 자세 추정을 어렵게 한다. 이를 해결하기 위해 본 논문에서는 딥러닝 기반 이상치 검출 및 보정 방법을 제안한다. 딥러닝 기반의 이상치 검출 방법은 여러 모션에서 좌표의 이상치를 효과적으로 검출하고, 모션의 특징을 활용한 규칙 기반 보정 방법을 통해 이상치를 보정한다. 다양한 실험과 분석을 통하여 제안하는 방법이 골프 스윙 모션과 다양한 운동 모션에서도 사람의 자세를 정확히 추정할 수 있고, 3차원 좌표 데이터에서도 확장 가능함을 보인다.

• 스마트미디어저널
• /
• 제13권4호
• /
• pp.16-22
• /
• 2024

투구는 야구의 시작이라 할 만큼 야구에서 주요한 부분을 차지한다. 투구 동작의 정확한 분석은 경기력 향상과 부상 예방 측면에서 매우 중요하다. 올바른 투구 동작을 분석할 때, 현재 주로 사용되는 모션캡처는 환경적으로 치명적인 단점들이 몇 가지 존재한다. 본 논문에서 우리는 이러한 단점들이 존재하는 모션캡처를 대체하기 위하여 RGB 기반의 Human Pose Estimation(HPE) 모델을 활용한 투구 동작의 분석을 제안하며 이에 대한 신뢰도를 검증하기 위해 모션캡처 데이터와 HPE 데이터의 관절 좌표를 Dynamic Time Warping(DTW) 알고리즘의 비교를 통해 두 데이터의 유사도를 검증하였다.

• 전기학회논문지
• /
• 제64권9호
• /
• pp.1347-1355
• /
• 2015

In this study, we compare the recognition performance using PCA and RBFNNs for introducing robust face recognition system to pose variations based on pose estimation. proposed face recognition system uses Honda/UCSD database for comparing recognition performance. Honda/UCSD database consists of 20 people, with 5 poses per person for a total of 500 face images. Extracted image consists of 5 poses using Multiple-Space PCA and each pose is performed by using (2D)²PCA for performing pose classification. Linear polynomial function is used as connection weight of RBFNNs Pattern Classifier and parameter coefficient is set by using Particle Swarm Optimization for model optimization. Proposed (2D)²PCA-based face pose classification performs recognition performance with PCA, (2D)²PCA and RBFNNs.