Pedestrian recognition in multi-camera networks based on deep transfer learning and feature visualization

Wang, Jingtao; Yan, Guoli; Wang, Huiyan; Hua, Jing

doi:10.1016/j.neucom.2018.07.063

Cited by 14 publications

(5 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Rudimentary models, while offering acceptable performance in classification of test data, have not been trained on a dataset as large as those which make use of transfer learning. Comparing the results in [43] where they used three different datasets and trained models using transfer learning, the results in this paper are very similar. The authors achieved an accuracy of 96.71% with 2000 training samples and 99.52% with 5000 training samples using SVM classifiers on the PRID database.…”

Section: Transfer Learning Adaptive Modelsupporting

confidence: 54%

Deep and Transfer Learning Approaches for Pedestrian Identification and Classification in Autonomous Vehicles

2021

View full text Add to dashboard Cite

Pedestrian detection is at the core of autonomous road vehicle navigation systems as they allow a vehicle to understand where potential hazards lie in the surrounding area and enable it to act in such a way that avoids traffic-accidents, which may result in individuals being harmed. In this work, a review of the convolutional neural networks (CNN) to tackle pedestrian detection is presented. We further present models based on CNN and transfer learning. The CNN model with the VGG-16 architecture is further optimised using the transfer learning approach. This paper demonstrates that the use of image augmentation on training data can yield varying results. In addition, a pre-processing system that can be used to prepare 3D spatial data obtained via LiDAR sensors is proposed. This pre-processing system is able to identify candidate regions that can be put forward for classification, whether that be 3D classification or a combination of 2D and 3D classifications via sensor fusion. We proposed a number of models based on transfer learning and convolutional neural networks and achieved over 98% accuracy with the adaptive transfer learning model.

show abstract

Section: Transfer Learning Adaptive Modelsupporting

confidence: 54%

Deep and Transfer Learning Approaches for Pedestrian Identification and Classification in Autonomous Vehicles

2021

View full text Add to dashboard Cite

show abstract

“…The structure vector from the nose to the left knee is represented by The vector angle is chosen as a motion feature to represent the various motion states of the human body because when people move, the angle of their limbs will change significantly. The cosine values of the clip angles in each of the four structural vector groups mentioned above are computed independently and serve as characteristics of the fall behaviour [18]. The cosine values of the two sets of constructed vectors' angle cosines are described as follows.：…”

Section: Feature Extraction Of Trajectoriesmentioning

confidence: 99%

Recognising pedestrian behaviour using a multi-channel spatiotemporal fusion network

Junnian Wang, Jianxin Li, Chengtong Zhuang

2024

jes

View full text Add to dashboard Cite

To protect public safety, video abnormal behaviour detection is essential. The low accuracy of identifying abnormal behaviour in pedestrians is the focus of this study. To address this, an abnormal target identification method based on multi-feature fusion of trajectory skeleton is proposed. The process begins with defining the type of abnormal behaviour in accordance with the environmental requirements. Next, pedestrian identification is carried out in the designated area, the pedestrian is tracked to determine its movement trajectory, and the image coordinates of the corresponding relevant nodes are calculated by analysing the human posture. Lastly, the track and skeleton features are integrated to classify the normal and abnormal behaviours, and the identification of abnormal target behaviours is accomplished. Test results on the behavioural analysis database indicate that this algorithm's accuracy is 87.08%, which is higher than the single-feature identification method's detection efficiency of abnormal behaviours.

show abstract

“…Generally, the research process of note starting point detection is summarized into three parts: preprocessing, signal reduction, and peak extraction. Wang et al used AdaBoost algorithm to select a set of features from the audio feature set and then classified the music genres by aggregation algorithm [7]. Engin et al used modulation spectrum analysis to capture time-varying information and prosody information in music signals [8].…”

Section: Mnfr-related Research Nazemi Et Al Comprehensivelymentioning

confidence: 99%

Musical Note Feature Recognition Based on Long Short-Term Memory

Yuan

2022

Mobile Information Systems

View full text Add to dashboard Cite

The timbre and volume of a single tone are among its fundamental characteristics. The single-tone detection technology is the key to the foundation of MNFR (musical note feature recognition), which is built on the fundamental feature extraction of single tones. A MNFR method based on LSTM (long short-term memory) is proposed because traditional methods have low accuracy in note feature classification and low accuracy in MNFR. To process the series of convolution feature maps, the feature maps are directly input into LSTM to learn hash codes. Extract the note features. Segment according to the changing trend of the physical features of the notes. Additionally, a number of feature maps are built from convolution feature maps extracted from numerous convolution layers of previously trained CNNs (convolutional neural networks), taking into account the spatial specifics and semantic features. The note start vector is produced using the enhanced peak extraction algorithm based on Gaussian kernel smoothing. The findings indicate that, when 100 samples are used, this method’s note classification accuracy differs from that of the DBN (deep belief network) and DWT (discrete wavelet transform) by 1.17 percent and 2.04 percent, respectively. The analysis in the conclusion demonstrates that the algorithm put forth in this paper is both theoretically and practically workable.

show abstract

Pedestrian recognition in multi-camera networks based on deep transfer learning and feature visualization

Cited by 14 publications

References 15 publications

Deep and Transfer Learning Approaches for Pedestrian Identification and Classification in Autonomous Vehicles

Deep and Transfer Learning Approaches for Pedestrian Identification and Classification in Autonomous Vehicles

Recognising pedestrian behaviour using a multi-channel spatiotemporal fusion network

Musical Note Feature Recognition Based on Long Short-Term Memory

Contact Info

Product

Resources

About