Robust 3D Face Landmark Localization Based on Local Coordinate Coding

Song, Mingli; Tao, Dacheng; Sun, Shengpeng; Chen, Chun; Maybank, Stephen J.

doi:10.1109/tip.2014.2361204

Cited by 20 publications

(4 citation statements)

References 34 publications

(41 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Besides, dual dictionary learning has been successfully applied to biomedical imaging as well [14–16]. Recently, local coordinate coding (LCC) based dictionary learning, which can find non-zero coefficients for dictionary atoms that are neighbors of the target sample for SR, has shown promising results in many applications [17–21]. Inspired by LCC, we propose an efficient semi-supervised tripled dictionary learning method to predict the S-PET image from L-PET and multimodal MRI.…”

Section: Introductionmentioning

confidence: 99%

Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI

Wang

et al. 2017

IEEE Trans. Biomed. Eng.

View full text Add to dashboard Cite

Objective To obtain high-quality positron emission tomography (PET) image with low-dose tracer injection, this study attempts to predict the standard-dose PET (S-PET) image from both its low-dose PET (L-PET) counterpart and corresponding magnetic resonance imaging (MRI). Methods It was achieved by patch-based sparse representation (SR), using the training samples with a complete set of MRI, L-PET and S-PET modalities for dictionary construction. However, the number of training samples with complete modalities is often limited. In practice, many samples generally have incomplete modalities (i.e., with one or two missing modalities) that thus cannot be used in the prediction process. In light of this, we develop a semi-supervised tripled dictionary learning (SSTDL) method for S-PET image prediction, which can utilize not only the samples with complete modalities (called complete samples) but also the samples with incomplete modalities (called incomplete samples), to take advantage of the large number of available training samples and thus further improve the prediction performance. Results Validation was done on a real human brain dataset consisting of 18 subjects, and the results show that our method is superior to the SR and other baseline methods. Conclusion This work proposed a new S-PET prediction method, which can significantly improve the PET image quality with low-dose injection. Significance The proposed method is favorable in clinical application since it can decrease the potential radiation risk for patients.

show abstract

Section: Introductionmentioning

confidence: 99%

Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI

Wang

et al. 2017

IEEE Trans. Biomed. Eng.

View full text Add to dashboard Cite

show abstract

“…In [5], the feature representations of both 2D and 3D samples were learned by radial basis function (RBF) networks in an unsupervised way, then the feature representation of a given image was reconstructed using the sample images' features and the reconstruction weights were used to obtain the personalised 3D feature representation from which the 3D face could be reconstructed through the learned RBF network with respect to the 3D samples. In [29], the authors used similar coupled RBF (CRBF) networks for 3D facial expression reconstruction, and then enhanced the expressions by adding expression details learned through the combination of a coupled dictionary and local coordinate coding [30]. Song et al [31] presented joint sparse learning to learn mapping functions and their respective inverses to model the relationship between the highdimensional 3D faces and their corresponding low-dimensional representations.…”

Section: Methods Based On Machine Learningmentioning

confidence: 99%

Approach to 3D face reconstruction through local deep feature alignment

Zhang

Zhu

2019

IET Computer Vision

View full text Add to dashboard Cite

Here, the authors propose an end-to-end method based on deep learning to reconstruct three-dimensional (3D) face models from given face images. In the training stage, the authors propose to extract the feature representations from the 3D sample faces and corresponding 2D sample images through the proposed local deep feature alignment (LDFA) algorithm, and estimate an explicit mapping from the 2D features to their 3D counterparts for each local neighbourhood, then the authors learn a feed-forward deep neural network for each neighbourhood whose parameters are initialised with the parameters obtained in the locality-aware learning process and the explicit mapping. In the testing stage, the authors only need to feed a given face image to the deep neural network corresponding to the nearest sample image and receive the outputted 3D face model. Extensive experiments have been conducted on both non-face and face data sets. The authors find that the LDFA algorithm performs better than several popular unsupervised feature extraction algorithms, and the 3D reconstruction results obtained by the proposed method also outperform the comparison methods.

show abstract

“…Similarly in our approach the algorithm takes in an RGB image and after performing various steps of skin tone segmentation [16] as discussed above followed by several morphological and filtering processes, it outputs the same RGB image with the largest possible face-like Blob surrounded by a rectangular bounding box [17] as the detected face. When these processes continue iteratively in a loop, the system tracks the user's face by locating the coordinates of the face-like blob within the preview.…”

Section: ) Real-time Face Trackingmentioning

confidence: 99%

A Real-Time Face Motion Based Approach towards Modeling Socially Assistive Wireless Robot Control with Voice Recognition

Bhattacharjee¹,

Das²,

Kundu³

et al. 2015

ijacsa

View full text Add to dashboard Cite

Abstract-The robotics domain has a couple of specific general design requirements which requires the close integration of planning, sensing, control and modeling and for sure the robot must take into account the interactions between itself, its task and its environment surrounding it. Thus considering the fundamental configurations, the main motive is to design a system with user-friendly interfaces that possess the ability to control embedded robotic systems by natural means. While earlier works have focused primarily on issues such as manipulation and navigation only, this proposal presents a conceptual and intuitive approach towards man-machine interaction in order to provide a secured live biometric logical authorization to the user access, while making an intelligent interaction with the control station to navigate advanced gesture controlled wireless Robotic prototypes or mobile surveillance systems along desired directions through required displacements. The intuitions are based on tracking real-time 3-Dimensional Face Motions using skin tone segmentation and maximum area considerations of segmented face-like blobs, Or directing the system with voice commands using real-time speech recognition. The system implementation requires designing a user interface to communicate between the Control station and prototypes wirelessly, either by accessing the internet over an encrypted WiFi Protected Access (WPA) via a HTML web page for communicating with face motions or with the help of natural voice commands like "Trace 5 squares", "Trace 10 triangles", "Move 10 meters", etc. evaluated on an iRobot Create over Bluetooth connectivity using a Bluetooth Access Module (BAM). Such an implementation can prove to be highly effective for designing systems of elderly aid and maneuvering the physically challenged.

show abstract

Robust 3D Face Landmark Localization Based on Local Coordinate Coding

Cited by 20 publications

References 34 publications

Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI

Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI

Approach to 3D face reconstruction through local deep feature alignment

A Real-Time Face Motion Based Approach towards Modeling Socially Assistive Wireless Robot Control with Voice Recognition

Contact Info

Product

Resources

About