scholarly journals Detail 3D Face Reconstruction Based on 3DMM and Displacement Map

2021 ◽  
Vol 2021 ◽  
pp. 1-13
Author(s):  
Tianping Li ◽  
Hongxin Xu ◽  
Hua Zhang ◽  
Honglin Wan
Keyword(s):  
Texture Features ◽  
Facial Feature ◽  
Reconstruction Method ◽  
3D Face Reconstruction ◽  
Face Surface ◽  
Face Model ◽  
Face Reconstruction ◽  
Adversarial Network ◽  
Face Features ◽  
The Face

How to accurately reconstruct the 3D model human face is a challenge issue in the computer vision. Due to the complexity of face reconstruction and diversity of face features, most existing methods are aimed at reconstructing a smooth face model with ignoring face details. In this paper a novel deep learning-based face reconstruction method is proposed. It contains two modules: initial face reconstruction and face details synthesis. In the initial face reconstruction module, a neural network is used to detect the facial feature points and the angle of the pose face, and 3D Morphable Model (3DMM) is used to reconstruct the rough shape of the face model. In the face detail synthesis module, Conditional Generation Adversarial Network (CGAN) is used to synthesize the displacement map. The map provides texture features to render to the face surface reconstruction, so as to reflect the face details. Our proposal is evaluated by Facescape dataset in experiments and achieved better performance than other current methods.

scholarly journals Fast Modelling Algorithm for Realistic Three-Dimensional Human Face for Film and Television Animation

Complexity ◽  
2021 ◽  
Vol 2021 ◽  
pp. 1-10
Author(s):  
Limin Xu
Keyword(s):  
Three Dimensional ◽  
Reconstruction Method ◽  
Feature Points ◽  
Human Face ◽  
3D Face Reconstruction ◽  
3D Face ◽  
Film And Television ◽  
Face Reconstruction ◽  
Reconstruction Performance ◽  
The Face

Aiming at the face photos of film and television animation, this paper proposes a new fast three-dimensional (3D) face modelling algorithm. First of all, based on the LBF algorithm, this paper proposes a multifeature selection idea to automatically detect multiple features of the face. Secondly, in order to solve the shortcomings of slow training speed while achieving large pose face alignment, the regression-based CNN is selected as the algorithm to achieve fast convergence. Then, due to the influence of various factors, the extracted feature points are not completely correct, and Gabor features are used to screen the matching of feature points. Finally, by analysing the principle of 3DMM 3D face reconstruction, a single-view 3D face reconstruction method based on CNN is proposed. The experimental results show that the algorithm in this paper has good reconstruction performance and real-time performance and can realize the rapid modelling of human face.

scholarly journals HifiFace: 3D Shape and Semantic Prior Guided High Fidelity Face Swapping

2021 ◽  
Author(s):  
Yuhan Wang ◽  
Xu Chen ◽  
Junwei Zhu ◽  
Wenqing Chu ◽  
Ying Tai ◽  
...  
Keyword(s):  
State Of The Art ◽  
High Fidelity ◽  
Reconstruction Method ◽  
3D Face Reconstruction ◽  
3D Shape ◽  
Face Shape ◽  
Face Reconstruction ◽  
The Face ◽  
In The Wild ◽  
Previous State

In this work, we propose a high fidelity face swapping method, called HifiFace, which can well preserve the face shape of the source face and generate photo-realistic results. Unlike other existing face swapping works that only use face recognition model to keep the identity similarity, we propose 3D shape-aware identity to control the face shape with the geometric supervision from 3DMM and 3D face reconstruction method. Meanwhile, we introduce the Semantic Facial Fusion module to optimize the combination of encoder and decoder features and make adaptive blending, which makes the results more photo-realistic. Extensive experiments on faces in the wild demonstrate that our method can preserve better identity, especially on the face shape, and can generate more photo-realistic results than previous state-of-the-art methods. Code is available at: https://johann.wang/HifiFace

3D Face Reconstruction from Two Orthogonal Images for Face Recognition Applications

2012 ◽  
pp. 223-239
Author(s):  
Stefano Berretti ◽  
Alberto Del Bimbo ◽  
Pietro Pala
Keyword(s):  
Face Recognition ◽  
Control Points ◽  
3D Face Reconstruction ◽  
3D Face ◽  
Face Reconstruction ◽  
Face Images ◽  
Model Control ◽  
3D Geometry ◽  
The Face ◽  
Manual Assistance

In this paper, an original hybrid 2D-3D face recognition approach is proposed using two orthogonal face images, frontal and side views of the face, to reconstruct the complete 3D geometry of the face. This is obtained using a model based solution, in which a 3D template face model is morphed according to the correspondence of a limited set of control points identified on the frontal and side images in addition to the model. Control points identification is driven by an Active Shape Model applied to the frontal image, whereas subsequent manual assistance is required for control points localization on the side view. The reconstructed 3D model is finally matched, using the iso-geodesic regions approach against a gallery of 3D face scans for the purpose of face recognition. Preliminary experimental results are provided on a small database showing the viability of the approach.

scholarly journals Enhancing the feature-based 3D deformable face recognition using hybrid PCA-NN

2021 ◽  
Vol 22 (1) ◽  
pp. 215
Author(s):  
Cahyo Darujati ◽  
Supeno Mardi Susiki Nugroho ◽  
Deny Kurniawan ◽  
Mochamad Hariadi
Keyword(s):  
Graphical Model ◽  
Original Learning ◽  
Principal Component ◽  
Facial Feature ◽  
Theory Approach ◽  
Face Features ◽  
The Neural Network ◽  
The Face ◽  
Critical Components ◽  
Almost All

<p>Facial recognition is one of the most important advancements in image processing. An important job is to build an automated framework with the same human capacity’s for recognizing face. The face is a complex 3D graphical model, and constructing a computational model is a challenging task. This paper aims at a facial detection technique focused on the coding and decoding of the facial feature object theory approach to data. One of the most natural and common principal component analysis (PCA) method. This approach transforms the face features into a minimal set of basic attributes, peculiarities, which are the critical components of the original learning image collection (or the training package). The proposed technique is a combination of the PCA system and the identification of components using the neural network (NN) feed-forward propagation method. This experiment proves that recognition of deformed 3D face is doable. By taking into account almost all forms of feature extraction and engineering, the NN yields a recognition score of 95%.</p>

scholarly journals 3D Face Reconstruction Techniques: Passive Methods

2019 ◽  
Vol 8 (2) ◽  
pp. 4354-4364
Keyword(s):  
3D Face Reconstruction ◽  
3D Images ◽  
3D Face ◽  
Image Reproduction ◽  
Face Reconstruction ◽  
The Face ◽  
History Of ◽  
Human Pose ◽  
Improved Accuracy ◽  
Pose Identification

In the recent literature, 3D face reconstruction received wide interest and has become one of the significant areas of research. 3D face reconstruction provides in depth details on geometrics, texture and color of the face, which are utilized in different applications. It supports a multitude of applications, ranging from face recognition and surveillance to medical imaging, gaming, animation, and virtual reality. This paper attempts to consolidate the research works that have happened in the history of 3D face reconstruction. Also, we try to classify the existing methods based on the input for the process. The databases used in the recent works are discussed and the performance evaluation of methods on different databases is analyzed. The challenges addressed in recent studies are mainly focused on the faster reconstruction of 3D Images, improved accuracy of reconstructed images, human pose identification, image reproduction with higher resolution. Researchers have also tried to address occlusion related problems. Passive methods, used by different researchers are analyzed and their effects on different parameters are discussed in this work. Finally, possible future areas for improvement in terms of reconstructions are presented for the benefit of researchers.

scholarly journals Application of Multiscale Facial Feature Manifold Learning Based on VGG-16

2021 ◽  
Vol 2021 ◽  
pp. 1-7
Author(s):  
Huilin Ge ◽  
Zhiyu Zhu ◽  
Runbang Liu ◽  
Xuedong Wu
Keyword(s):  
Face Detection ◽  
Recognition Rate ◽  
Facial Feature ◽  
Time Dimension ◽  
Running Time ◽  
The Public ◽  
Detection And Tracking ◽  
Face Features ◽  
Face Target ◽  
The Face

Purpose. In order to solve the problems of small face image samples, high size, low structure, no label, and difficulty in tracking and recapture in security videos, we propose a popular multiscale facial feature manifold (MSFFM) algorithm based on VGG16. Method. We first build the VGG16 architecture to obtain face features at different scales and construct a multiscale face feature manifold with face features at different scales as dimensions. At the same time, the recognition rate, accuracy rate, and running time are used to evaluate the performance of VGG16, LeNet-5, and DenseNet on the same database. Results. From the results of comparative experiments, it can be seen that the recognition rate and accuracy of VGG16 are the highest among the three networks. The recognition rate of VGG16 is 97.588%, and the accuracy is 95.889%. And the running time is only 3.5 seconds, which is 72.727% faster than LeNet-5 and 66.666% faster than DenseNet. Conclusion. The model proposed in this paper breaks through the key problem in the face detection and tracking problem in the public security field, predicts the position of the face target image in the time dimension manifold space, and improves the efficiency of face detection.

scholarly journals A Robust Shape Reconstruction Method for Facial Feature Point Detection

2017 ◽  
Vol 2017 ◽  
pp. 1-11
Author(s):  
Shuqiu Tan ◽  
Dongyi Chen ◽  
Chenggang Guo ◽  
Zhiqi Huang
Keyword(s):  
Feature Space ◽  
Facial Feature ◽  
Shape Space ◽  
Shape Reconstruction ◽  
Feature Point ◽  
Reconstruction Method ◽  
Large Variability ◽  
The Face ◽  
Public Datasets ◽  
Point Detection

Facial feature point detection has been receiving great research advances in recent years. Numerous methods have been developed and applied in practical face analysis systems. However, it is still a quite challenging task because of the large variability in expression and gestures and the existence of occlusions in real-world photo shoot. In this paper, we present a robust sparse reconstruction method for the face alignment problems. Instead of a direct regression between the feature space and the shape space, the concept of shape increment reconstruction is introduced. Moreover, a set of coupled overcomplete dictionaries termed the shape increment dictionary and the local appearance dictionary are learned in a regressive manner to select robust features and fit shape increments. Additionally, to make the learned model more generalized, we select the best matched parameter set through extensive validation tests. Experimental results on three public datasets demonstrate that the proposed method achieves a better robustness over the state-of-the-art methods.

Sign in / Sign up

Export Citation Format

Share Document