Occlusion Handling and Multi-Scale Pedestrian Detection Based on Deep Learning: A Review

Fang, Li; Li, Xueyuan; Liu, Qi; Li, Zirui

doi:10.1109/access.2022.3150988

Cited by 24 publications

(11 citation statements)

References 115 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Spinning radar focuses on CNNs to encode the representation. Similar to the occlusion issue in computer vision, most models try to generate a mask or weight function to deal with the noisy and potentially faulty radar data [145].…”

Section: Discussion and Future Researchmentioning

confidence: 99%

“…Although state-of-the-art deep learning is used, the number of papers published for radar odometry and localization is far below the number of pages published for object detection and segmentation for radars [5,9]. Similar to [107], the results from segmentation papers [145][146][147] can be used to mask out irrelevant detections and areas from scans that do not contribute to the odometry or localization of the system. In a complete pipeline, segmentation, classification, odometry, and ego-localization should be performed in an end-to-end approach, where the results from one module are integrated, e.g.…”

Section: Discussion and Future Researchmentioning

confidence: 99%

See 1 more Smart Citation

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Brune,

Meisen,

Pomp

2024

Applied Sciences

View full text Add to dashboard Cite

This paper provides an in-depth review of deep learning techniques to address the challenges of odometry and global ego-localization using frequency modulated continuous wave (FMCW) radar sensors. In particular, we focus on the prediction of odometry, which involves the determination of the ego-motion of a system by external sensors, and loop closure detection, which concentrates on the determination of the ego-position typically on an existing map. We initially emphasize the significance of these tasks in the context of radar sensors and underscore the motivations behind them. The subsequent sections delve into the practical implementation of deep learning approaches, strategically designed to effectively address the aforementioned challenges. We primarily focus on spinning and automotive radar configurations within the domain of autonomous driving. Additionally, we introduce publicly available datasets that have been instrumental in addressing these challenges and analyze the importance and struggles of current methods used for radar based odometry and localization. In conclusion, this paper highlights the distinctions between the addressed tasks and other radar perception applications, while also discussing their differences from challenges posed by alternative sensor modalities. The findings contribute to the ongoing discourse on advancing radar sensor capabilities through the application of deep learning methodologies, particularly in the context of enhancing odometry and ego-localization for autonomous driving applications.

show abstract

Section: Discussion and Future Researchmentioning

confidence: 99%

Section: Discussion and Future Researchmentioning

confidence: 99%

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Brune,

Meisen,

Pomp

2024

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…The proposed technique was used to compare the social force model of [58], sparse representation of [55], and mixture of dynamic textures method of [40], [59], [60]. The performance is assessed using standard performance evaluation metrics.…”

Section: Suspicious Activity Recognitionmentioning

confidence: 99%

Novel Pedestrian Detection and Suspicious Activity Recognition Using Enhanced YOLOv5 and Motion Feature Map

Gawande¹,

Hajari²,

Golhar

2023

Preprint

View full text Add to dashboard Cite

Pedestrian detection and suspicious activity recognition are notable challenges in vision-based surveillance systems. However, the accuracy of pedestrian detection is influenced by a wide range of factors, including human presence, trajectory, posture, complex background, and object deformation. In this paper, we developed a pedestrian dataset that encompasses student behavior in an institution, such as cheating, stealing lab devices, dispute, and threatening scenarios. It provides uniform and steady identification annotations of pedestrians, making it suitable for pedestrian detection, tracking, and behavior detection. In addition, we also improved detection accuracy using the enhanced YOLOv5 architecture. Next, we proposed an efficient method for detection of global and local unusual behaviors. The proposed method extracts motion features that accurately describe pedestrian motion, speed, and direction, etc. The proposed method is validated on a proposed student behavior database and three additional publicly available benchmark datasets. It gives state-of-the-art detection accuracy of 96.12% and miss rate of 6.68% compared to the existing techniques. The experimental results demonstrate a substantial performance improvement in anomalous activity recognition. The final section summarizes and discusses future research directions.

show abstract

“…In addition, when the detection confidence score is less than the threshold (ID = 1 at t1) or the worker is in unusual behavior/posture variation [11] (ID = 2 at t3), the bounding box may be missed. Previous studies' successes were owing to the development of more complicated networks or having extensively re-trained on new datasets; however, their effectiveness in addressing the challenges of complex scenarios was limited [12] by having to manually label new datasets [13], re-train entire neural networks [14,15], or switch to offline features [16]. Regardless of the chosen approaches, they primarily relied on the convolutional neural network (CNN) [10] outputs.…”

Section: Introductionmentioning

confidence: 99%

Head-Integrated Detecting Method for Workers under Complex Construction Scenarios

Liu,

Zhou,

Wang

et al. 2024

Buildings

View full text Add to dashboard Cite

Real-time detection of workers is crucial in construction safety management. Deep learning-based detecting methods are valuable, but always challenged by the possibility of target missing or identity errors under complex scenarios. To address these limitations, previous research depended on re-training for new models or datasets, which are prohibitively time-consuming and incur high computing demands. However, we demonstrate that the better detecting model might not rely on more re-training of weights; instead, a training-free model can achieve even better performance by integrating head information. In this paper, a new head-detecting branch (55 MB) is added to the Keypoint Region-based Convolutional Network (Keypoint R-CNN, 226 MB) without altering its original weights, allowing for a less occluded head to aid in body detection. We also deployed motion information and anthropometric data through a post-processing module to calculate movement relationships. This study achieved an identity F1-score (IDF1) of 97.609%, recall (Rcll) of 98.173%, precision (Prcn) of 97.052%, and accuracy of 95.329% as a state-of-the-art (SOTA) method for worker detection. This exploration breaks the inertial attitudes of re-training dependency and accelerates the application of universal models, in addition to reducing the computational difficulty for most construction sites, especially in scenarios with an insufficient graphics processing unit (GPU). More importantly, this study can address occlusion challenges effectively in the worker detection field, making it of practical significance.

show abstract

Occlusion Handling and Multi-Scale Pedestrian Detection Based on Deep Learning: A Review

Cited by 24 publications

References 115 publications

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Survey of Deep Learning-Based Methods for FMCW Radar Odometry and Ego-Localization

Novel Pedestrian Detection and Suspicious Activity Recognition Using Enhanced YOLOv5 and Motion Feature Map

Head-Integrated Detecting Method for Workers under Complex Construction Scenarios

Contact Info

Product

Resources

About