Learning rich features from objectness estimation for human lying-pose detection

Xia, Daoxun; Su, Songzhi; Lu, Geng; Wu, Gang; Li, Shaozi

doi:10.1007/s00530-016-0518-5

Cited by 10 publications

(6 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Batch Normalization algorithm is one of the most exciting innovations in optimizing deep convolutional neural networks. It is a data normalization algorithm proposed by literature [30] to alleviate the covariance migration phenomenon.…”

Section: Optimization Methods Of Convolutional Neural Networkmentioning

confidence: 99%

Brain Image Recognition Algorithm and High Performance Computing of Internet of Medical Things Based on Convolutional Neural Network

Liu¹,

Xiong

2019

IEEE Access

View full text Add to dashboard Cite

Due to the wide variety of medical images and the complexity of the human body structure, the characteristics of manual extraction of medical images are difficult, the adaptive ability is poor, and the classification effect needs to be improved. Aiming at the shortcomings of traditional medical image recognition methods, this paper proposes an adaptive convolutional neural network model CNN-BN-PReLU based on the convolutional neural network method. The model first performs batch normalization (BN) processing on the input of each feature map of each layer of network, and then adaptively adjusts the parameters by using Parametric Rectified Linear Unit (PReLU) to compare the BN algorithm. Based on the performance before and after the activation function, an adaptive convolutional neural network model is constructed. The experimental results show that the model can abstract the image features without artificial intervention, speed up the network convergence speed and shorten the training time, and significantly improve the image recognition rate and reduce the misdiagnosis rate and missed diagnosis rate of the disease. INDEX TERMS Brain image recognition, high performance computing, convolutional neural network.

show abstract

Section: Optimization Methods Of Convolutional Neural Networkmentioning

confidence: 99%

Brain Image Recognition Algorithm and High Performance Computing of Internet of Medical Things Based on Convolutional Neural Network

Liu¹,

Xiong

2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Lying pose detection is more challenging than basic human (pedestrian) detection [13], [14]. Bodies lying on the ground may have arbitrary positions and configurations and may suffer from severe perspective distortions.…”

Section: B Lying Pose Detectionmentioning

confidence: 99%

“…Lying pose detection has important uses in numerous applications [14]. One such application is fall detection for elders and persons with disabilities living in smart homes [5], [7], [22].…”

Section: Introductionmentioning

confidence: 99%

Automatic Dataset Expansion With Structured Feature Learning for Human Lying Pose Detection

et al. 2020

View full text Add to dashboard Cite

In this study, we developed a framework to localize human lying poses by a camera positioned above. Our framework is motivated by the fact that detecting lying poses is fundamentally more difficult than detecting pedestrians or localizing nondeformable objects such as cars, roads, and buildings due to the large number of poses, orientations, and scales that a human lying on the ground can take. An important problem with lying pose detection is the training dataset, which hardly accounts for each possible body configuration. As a solution, we propose a geometric expansion procedure that uses a virtual camera to increase the number of training images. We also use a Gibbs sampler to generate more training samples in the feature space on which the system can train its model. Once the training is completed, detection is performed on a multiscale and multirotational space. Because our framework accommodates a variety of object detection systems, we report the results for the Faster R-CNN, FPN, and RefineDet models. The results show that using automatic dataset expansion models systematically improves the results. INDEX TERMS Human lying pose detection, automatic dataset expansion, perspective transformation, gibbs sampling, deep learning.

show abstract

“…where the authors in [6] presented two methods for human body pose recognition based on deep learning using a convolutional neural network (CNN) and optimized neural network with similarity evaluation between dimensions for mapping human body poses to the right estimation. Human body pose recognition task solving by deep learning in [7] using binarized normed gradient features to extract objectiveness using saliency map with CNN that learnt the hierarchies of the features and made predictions based on learnt.…”

Section: Related Workmentioning

confidence: 99%

Human Body Poses Recognition Using Neural Networks with Data Augmentation

Al-Qerem¹

2019

IJATCSE

View full text Add to dashboard Cite

The continuous development of technology of IT enabled computers to see and learn. There are many viable applications for computer learning and vision to solve new tasks. In this paper, we propose a framework, able of automatically perceiving the human body poses from a single image, acquired by a traditional low-cost camera. Our methodology exploiting the vision of computers features and neural networks to detect and recognize a human from an image. The processing start with detection human in image and then extracting the silhouette from an image then using a neural network to recognize body poses based on silhouettes that extracted. To classified the detected silhouettes with body poses, the neural network was trained with dataset of preprocessed images contains silhouettes and labelled using 0, 1 for standing and walking poses respectively labelled , according to our outcomes that resulted in an accuracy over 93% with the best neural network model. The proposed approach offers promising effects with accepted accuracy. This knowledge will help greatly in decision-making and provide independent vehicles with valuable information for their artificial intelligence to process and make decisions as regards the movement of pedestrians.

show abstract

Learning rich features from objectness estimation for human lying-pose detection

Cited by 10 publications

References 48 publications

Brain Image Recognition Algorithm and High Performance Computing of Internet of Medical Things Based on Convolutional Neural Network

Brain Image Recognition Algorithm and High Performance Computing of Internet of Medical Things Based on Convolutional Neural Network

Automatic Dataset Expansion With Structured Feature Learning for Human Lying Pose Detection

Human Body Poses Recognition Using Neural Networks with Data Augmentation

Contact Info

Product

Resources

About