Learning toward practical head pose estimation

Sang, Gaoli; He, Feifei; Zhu, Rong; Xuan, Shibin

doi:10.1117/1.oe.56.8.083104

Cited by 6 publications

(2 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, with the widespread use of deep neural networks, people have focused on realistic application requirements, combined with existing computer vision technologies, and designed algorithms for specific problems in different application scenarios and a large number of datasets and methods have been proposed one after another [8][9]. Among them, the behavioral assessment method based on pose estimation extracts the human skeleton data from ordinary RGB cameras without the help of additional hardware devices [10][11].…”

Section: Introductionmentioning

confidence: 99%

Human pose evaluation based on full-domain convolution and LSTM

Zou,

Pan,

Zhou

et al. 2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

In this paper, we first analyze full domain convolution and LSTM to evaluate human pose by convolutional neural network and LSTM network. Secondly, graph structure skeleton image and skeleton point image classifier based on CNN and LSTM is constructed. The two-dimensional pose assessment method and three-dimensional pose assessment method were used to empirically analyze the human pose assessment. The results show that the average accuracy mAP values of the traditional evaluation methods are 69.7, 72.3, 71.4, and 74.4, respectively, while the average accuracy mAP value of the method used for 2D pose evaluation is 74.6. Where the average error of LReLU is the smallest. This shows that full-domain convolution and LSTM can be effective for human pose evaluation.

show abstract

Section: Introductionmentioning

confidence: 99%

Human pose evaluation based on full-domain convolution and LSTM

Zou,

Pan,

Zhou

et al. 2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

show abstract

“…But practical applications prove these assumptions to be invalid. To overcome these limitations, a coping methodology is proposed in [15], which shows the use of the deep convolutional neural network (DCNN) to handle undersampling and uncertainty. The proposed method uses dense sampling intervals with multivariate labeling distributions (MLDs) to show input face image head pose angles.…”

Section: Introductionmentioning

confidence: 99%

Attention Span Prediction Using Head-Pose Estimation With Deep Neural Networks

Singh

Mohadikar

Gite³

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Automated human pose estimation is evolving as an exciting research area in human activity detection. It includes sophisticated applications such as malpractice detection in the examination, distracted driving, gesture detection, etc., and requires robust and reliable pose estimation techniques. These applications help to map the attention of the user with head pose estimation (HPE) metrics supported by emotion and gaze analysis. This paper solves the problem of attention score estimation with HPE. The proposed method ensures ease of implementation while addressing head pose estimation using 68 facial features. Further, to attain reliability and precision, head pose estimation has been implemented as a regression task. The coordinate pair angle method (CPAM) with deep neural network (DNN) regression and elastic net regression is carried out. The use of DNN ensures precision on low lighting, distorted or occluded images. CPAM methodology leverages facial landmark detection and angular difference to estimate head pose. Experimentation results showed that the proposed model could handle large datasets, real-time data processing, significant pose variations, partial occlusions, and diverse facial expressions with a mean absolute error (MAE) of 3° and less. The proposed system was evaluated on three standard databases: the 300W across large poses (300W-LP) dataset, annotated facial landmarks in the wild (AFLW2000) dataset, and the national institute of mental health child emotional faces picture set (NIMH-ChEFS) dataset. The results achieved are on par with recent state-of-the-art methodologies such as anisotropic angle distribution learning (AADL), joint head pose estimation and face alignment algorithm (JFA), rotation axis focused attention network (RAFA-Net), and propose an MAE ranging up to 6°. The paper could achieve remarkable results for attention span prediction using head pose estimation and for many possible future applications.

show abstract

Evolutionary correlation filtering based on pseudo-bacterial genetic algorithm for pose estimation of highly occluded targets

Picos

Orozco-Rosas

2020

Multimed Tools Appl

View full text Add to dashboard Cite

Learning toward practical head pose estimation

Cited by 6 publications

References 28 publications

Human pose evaluation based on full-domain convolution and LSTM

Human pose evaluation based on full-domain convolution and LSTM

Attention Span Prediction Using Head-Pose Estimation With Deep Neural Networks

Evolutionary correlation filtering based on pseudo-bacterial genetic algorithm for pose estimation of highly occluded targets

Contact Info

Product

Resources

About