Scale Driven Convolutional Neural Network Model for People Counting and Localization in Crowd Scenes

Basalamah, Saleh; Khan, Sultan Daud; Ullah, Habib

doi:10.1109/access.2019.2918650

Cited by 65 publications

(44 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, design idea and frame expressed in this paper has have certain theoretical guidance meaning and practical reference value for the research of brain science. We will drill down further and explore deep learning [49]- [51] in our follow-on works.…”

Section: Comparison With Existing Workmentioning

confidence: 99%

Classification Epileptic Seizures in EEG Using Time-Frequency Image and Block Texture Features

Sun

Chen

et al. 2020

IEEE Access

View full text Add to dashboard Cite

With the rapid development in technology, computer aided detection or diagnosis has become an indispensable part of the medical industry. Automatic detection of epileptic events is one of the important subjects that have aroused wide interest from more and more investigators. This paper proposes a new model in classification of multi-category electroencephalogram (EEG) signals using time-frequency image and block texture features. The one-dimensional EEG is first mapped to time-frequency domain by means of short-time Fourier transform (STFT), which is adapted to obtain a two-dimensional time-frequency image (2D-TFI). With the idea of multi-scale blocking, the obtained phase images and amplitude images are divided into several sub-blocks corresponding to different frequency ranges and time periods. Then the texture features are calculated to describe the behaviour of EEG signals. Particularly, a novel quadratic feature selection method based on kernel entropy component analysis (KECA) and Kruskal-Wallis test (KW) has been proposed for dimension reduction, by which the features that contained most distinctive information were provided. Eventually, the optimal KECA-based features are fed to support vector machine (SVM) for deciding the class of corresponding EEG. The proposed model is found to achieve at least 99.30% accuracy, 98.0% sensitivity and 100% specificity for each of the eight clinical problems. Our scheme is proven to be effective for seizure detection, which can help doctors optimize the diagnosis workflow, reduce workload and improve detection precision. INDEX TERMS Seizure, EEG, two-dimensional time-frequency image, multi-scale blocking, texture features, kernel entropy component analysis, quadratic feature selection.

show abstract

Section: Comparison With Existing Workmentioning

confidence: 99%

Classification Epileptic Seizures in EEG Using Time-Frequency Image and Block Texture Features

Sun

Chen

et al. 2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Recently, CNN based approaches [19][20][21][22] have shown their advantages in learning the crowd image feature mapping and the people/head detection for both crowd counting [23][24][25][26][27][28] and localization [26][27][28][29][30]. The Multi-column Convolutional Neural Network (MCNN) method is evaluated in [19] which contains three columns of different filters to extract feature of heads in different scales.…”

Section: Cnns For Crowd Countingmentioning

confidence: 99%

“…Besides, Idrees [32] introduced a deep CNN with composition loss method to satisfy counting, density map estimation, and localization. To handle the small/tiny objects that often appear in crowd counting scenes, Basalamah et al [27] used the scale-aware object proposal generated by perspective information which handled scale variations and makes the model (SD-CNN) able to detect human heads in both low density and high-density crowd images. Onoro et al [33] using the Hydra-CNN fuses the multi-scale information provided by heads to handle the crowd counting problems with significant variations in the scene.…”

Section: Cnns For Crowd Countingmentioning

confidence: 99%

“…Chen et al [38] also introduced the new framework of EGR and introduced a new metric for both errors in localization and counting. Nowadays, most researches [26][27][28]33] using both the density map and neural network detector for the localization task. Idrees et al [32] introduced the composition loss to do the counting, density map estimation and localization in congested scenes simultaneously.…”

Section: Cnns For Localizationmentioning

confidence: 99%

“…Idrees et al [32] introduced the composition loss to do the counting, density map estimation and localization in congested scenes simultaneously. Reference [27] devises a scale-aware head detector and using the response map to optimize the detector to make the test results more consistent with population distribution. Instead of the scale-awareness, Reference [26] approached the depth information by designed a depth-aware anchor to initialize the anchor and estimated the bounding box sizes of all heads that were utilized as the ground truth to train the RDNet.…”

Section: Cnns For Localizationmentioning

confidence: 99%

See 2 more Smart Citations

Counting Crowds with Perspective Distortion Correction via Adaptive Learning

Sun

Jin

et al. 2020

Sensors

View full text Add to dashboard Cite

The goal of crowd counting is to estimate the number of people in the image. Presently, use regression to count people number became a mainstream method. It is worth noting that, with the development of convolutional neural networks (CNN), methods that are based on CNN have become a research hotspot. It is a more interesting topic that how to locate the site of the person in the image than simply predicting the number of people in the image. The perspective transformation present is still a challenge, because perspective distortion will cause differences in the size of the crowd in the image. To devote perspective distortion and locate the site of the person more accuracy, we design a novel framework named Adaptive Learning Network (CAL). We use the VGG as the backbone. After each pooling layer is output, we collect the 1/2, 1/4, 1/8, and 1/16 features of the original image and combine them with the weights learned by an adaptive learning branch. The object of our adaptive learning branch is each image in the datasets. By combining the output features of different sizes of each image, the challenge of drastic changes in the size of the image crowd due to perspective transformation is reduced. We conducted experiments on four population counting data sets (i.e., ShanghaiTech Part A, ShanghaiTech Part B, UCF_CC_50 and UCF-QNRF), and the results show that our model has a good performance.

show abstract

Different Approaches to Background Subtraction and Object Tracking in Video Streams: A Review

Sivanantham,

Blessington Praveen P,

Kumar

2023

Contributions to Environmental Sciences &Amp; Innovative Business Technology

View full text Add to dashboard Cite

Scale Driven Convolutional Neural Network Model for People Counting and Localization in Crowd Scenes

Cited by 65 publications

References 39 publications

Classification Epileptic Seizures in EEG Using Time-Frequency Image and Block Texture Features

Classification Epileptic Seizures in EEG Using Time-Frequency Image and Block Texture Features

Counting Crowds with Perspective Distortion Correction via Adaptive Learning

Different Approaches to Background Subtraction and Object Tracking in Video Streams: A Review

Contact Info

Product

Resources

About