Task-Oriented Feature-Fused Network With Multivariate Dataset for Joint Face Analysis

Lin, Xuxin; Wan, Jun; Xie, Yiliang; Zhang, Shifeng; Lin, Chao; Liang, Yanyan; Guo, Guodong; Li, Stan Z.

doi:10.1109/tcyb.2019.2917049

Cited by 13 publications

(9 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Huang et al [25] propose an efficient greedy neural architecture search approach (GNAS) to automatically discover the optimal treelike architecture for PAR. Lin et al [11] develop a new multi-task network for facial attribute analysis. They design task-oriented feature-fused blocks, where each task learns effective feature combinations for classification.…”

Section: A Person Attribute Recognitionmentioning

confidence: 99%

See 1 more Smart Citation

Drop Loss for Person Attribute Recognition With Imbalanced Noisy-Labeled Samples

Yan

Xue

et al. 2023

IEEE Trans. Cybern.

View full text Add to dashboard Cite

Person attribute recognition (PAR) aims to simultaneously predict multiple attributes of a person. Existing deep learning-based PAR methods have achieved impressive performance. Unfortunately, these methods usually ignore the fact that different attributes have an imbalance in the number of noisylabeled samples in the PAR training datasets, thus leading to suboptimal performance. To address the above problem of imbalanced noisy-labeled samples, we propose a novel and effective loss called drop loss for PAR. In the drop loss, the attributes are treated differently in an easy-to-hard way. In particular, the noisy-labeled candidates, which are identified according to their gradient norms, are dropped with a higher drop rate for the harder attribute. Such a manner adaptively alleviates the adverse effect of imbalanced noisy-labeled samples on model learning. To illustrate the effectiveness of the proposed loss, we train a simple ResNet-50 model based on the drop loss and term it DropNet. Experimental results on two representative PAR tasks (including facial attribute recognition and pedestrian attribute recognition) demonstrate that the proposed DropNet achieves comparable or better performance in terms of both balanced accuracy and classification accuracy over several state-of-the-art PAR methods.

show abstract

Section: A Person Attribute Recognitionmentioning

confidence: 99%

“…Recently, a large number of deep learning-based PAR methods [7]- [11] have been developed and have shown promising performance. These methods typically formulate the problem of predicting attributes as the problem of designing proper deep neural networks on the training datasets.…”

Section: Introductionmentioning

confidence: 99%

Drop Loss for Person Attribute Recognition With Imbalanced Noisy-Labeled Samples

Yan

Xue

et al. 2023

IEEE Trans. Cybern.

View full text Add to dashboard Cite

show abstract

“…A person's blink frequency changes in response to their level of activity. The number of blinks increases when a person reads a certain phrase aloud or performs a visually given information exercise, whereas the number of blinks decreases when a person focuses on visual information or reads words quietly [41].…”

Section: Facial Landmarks For Eye Blink Detectionmentioning

confidence: 99%

Eye Aspect Ratio for Real-Time Drowsiness Detection to Improve Driver Safety

et al. 2022

View full text Add to dashboard Cite

Drowsiness is a major risk factor for road safety, contributing to serious injury, death, and economic loss on the road. Driving performance decreases because of increased drowsiness. In several different applications, such as facial movement analysis and driver safety, blink detection is an essential requirement that is used. The extremely rapid blink rate, on the other hand, makes automatic blink detection an extremely challenging task. This research paper presents a technique for identifying eye blinks in a video series recorded by a car dashboard camera in real time. The suggested technique determines the facial landmark positions for each video frame and then extracts the vertical distance between the eyelids from the facial landmark positions. The algorithm that has been proposed estimates the facial landmark positions, extracts a single scalar quantity by making use of Eye Aspect Ratio (EAR), and identifies the eye closeness in each frame. In the end, blinks are recognized by employing the modified EAR threshold value in conjunction with a pattern of EAR values in a relatively short period of time. Experimental evidence indicates that the greater the EAR threshold, the worse the AUC’s accuracy and performance. Further, 0.18 was determined to be the optimum EAR threshold in our research.

show abstract

“…The techniques used for feature estimation are of versatile nature for this trio like image processing based [4] to wavelets etc. Some commercial applications [108] and PCA [99] etc. are used for classification or retrieval.…”

Section: Techniques For Improved Global Traits Based Recognitionmentioning

confidence: 99%

Soft biometrics: a survey

Hassan

Izquierdo

Piatrik

2021

Multimed Tools Appl

View full text Add to dashboard Cite

The field of biometrics research encompasses the need to associate an identity to an individual based on the persons physiological or behaviour traits. While the use of intrusive techniques such as retina scans and finger print identification has resulted in highly accurate systems, the scalability of such systems in real-world applications such as surveillance and border security has been limited. As a branch of biometrics research, the origin of soft biometrics could be traced back to need for non-intrusive solutions for extracting physiological traits of a person. Following high number of research outcomes reported in the literature on soft biometrics, this paper aims to consolidate the scope of soft biometrics research across four thematic schemes (i) a detailed review of soft biometrics research data sets, their annotation strategies and building a largest novel collection of soft traits; (ii) the assessment of metrics that affect the performance of soft biometrics system; (iii) a comparative analysis on feature and modality level fusion reported in the literature for enhancing the system performance; and (iv) a performance analysis of hybrid soft biometrics recognition system using multi-scale criterion. The paper also presents a detailed analysis on the global traits associated to person identity such as gender, age and ethnicity. The contribution of the paper is to provide a comprehensive review of scientific literature, identify open challenges and offer insights on new research directions in the filed.

show abstract

Task-Oriented Feature-Fused Network With Multivariate Dataset for Joint Face Analysis

Cited by 13 publications

References 45 publications

Drop Loss for Person Attribute Recognition With Imbalanced Noisy-Labeled Samples

Drop Loss for Person Attribute Recognition With Imbalanced Noisy-Labeled Samples

Eye Aspect Ratio for Real-Time Drowsiness Detection to Improve Driver Safety

Soft biometrics: a survey

Contact Info

Product

Resources

About