An evaluation of deep learning based object detection strategies for threat object detection in baggage security imagery

Sangwan, Dhiraj; Jain, Deepak Kumar

doi:10.1016/j.patrec.2019.01.014

Cited by 79 publications

(16 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The mAP achieved for maturity stage classification using Resnet-101 was 81.5% which is greater than the 68% accuracy reported in another study [14]. A model was designed in [15] to perform threat object detection using faster R-CNN and YOLO. The performance is studied on 4 classes of threat objects: i) Gun, ii) Shuriken, iii) Razor-blade, iv) Knife.…”

Section: Introductionmentioning

confidence: 67%

Automated object detection of mechanical fasteners using faster region based convolutional neural networks

Karthikeyan

Subashini

2021

IJECE

View full text Add to dashboard Cite

Mechanical fasteners are widely used in manufacturing of hardware and mechanical components such as automobiles, turbine & power generation and industries. Object detection method play a vital role to make a smart system for the society. Internet of things (IoT) leads to automation based on sensors and actuators not enough to build the systems due to limitations of sensors. Computer vision is the one which makes IoT too much smarter using deep learning techniques. Object detection is used to detect, recognize and localize the object in an image or a real time video. In industry revolution, robot arm is used to fit the fasteners to the automobile components. This system will helps the robot to detect the object of fasteners such as screw and nails accordingly to fit to the vehicle moved in the assembly line. Faster R-CNN deep learning algorithm is used to train the custom dataset and object detection is used to detect the fasteners. Region based convolutional neural networks (Faster R-CNN) uses a region proposed network (RPN) network to train the model efficiently and also with the help of Region of Interest able to localize the screw and nails objects with a mean average precision of 0.72 percent leads to accuracy of 95 percent object detection

show abstract

Section: Introductionmentioning

confidence: 67%

Automated object detection of mechanical fasteners using faster region based convolutional neural networks

Karthikeyan

Subashini

2021

IJECE

View full text Add to dashboard Cite

show abstract

“…The comparison is reported in Table II, where it can be seen that the proposed framework outperforms its competitors by 5.35% in terms of recall and 2.68% in terms of F-score. Also, it is worth noting that the proposed framework although lags from YOLOv2 [27] by 0.988% in terms of accuracy, it beats YOLOv2 [27] by 4.61% in terms of F-score, which is more promising considering the fact that accuracy is vulnerable against the imbalanced data (especially for recognizing true negatives 'background' pixels) compared to the F-score. Apart from this, the proposed framework also outperformed state-of-the-art occlusion-aware dual tensor-shot detector (DTSD) framework [29] by 2.6% and Cascaded Structure Tensor framework [11] by 1.22% in terms of mAP scores, as evident from Table III.…”

Section: B Comparative Analysismentioning

confidence: 93%

“…In the second set of evaluations, we compared the proposed framework's detection performance to that of state-of-the-art baggage threat detection frameworks like CST [11], DTSD [29], and the YOLOv2 [27], and Tiny YOLO [27] based scheme proposed in [27]. The comparison is reported in Table II, where it can be seen that the proposed framework outperforms its competitors by 5.35% in terms of recall and 2.68% in terms of F-score.…”

Section: B Comparative Analysismentioning

confidence: 99%

Temporal Fusion Based Mutli-scale Semantic Segmentation for Detecting Concealed Baggage Threats

Shafay¹,

Hassan²,

Damiani³

et al. 2021

Preprint

View full text Add to dashboard Cite

Detection of illegal and threatening items in baggage is one of the utmost security concern nowadays. Even for experienced security personnel, manual detection is a timeconsuming and stressful task.Many academics have created automated frameworks for detecting suspicious and contraband data from X-ray scans of luggage. However, to our knowledge, no framework exists that utilizes temporal baggage X-ray imagery to effectively screen highly concealed and occluded objects which are barely visible even to the naked eye. To address this, we present a novel temporal fusion driven multiscale residual fashioned encoder-decoder that takes series of consecutive scans as input and fuses them to generate distinct feature representations of the suspicious and non-suspicious baggage content, leading towards a more accurate extraction of the contraband data. The proposed methodology has been thoroughly tested using the publicly accessible GDXray dataset, which is the only dataset containing temporally linked grayscale X-ray scans showcasing extremely concealed contraband data. The proposed framework outperforms its competitors on the GDXray dataset on various metrics.

show abstract

“…Apart from this, Xiao et al [70] presented an efficient implementation of Faster R-CNN [12] to detect suspicious data from the TeraHertz imagery. Dhiraj et al [71] used Faster R-CNN [12], YOLOv2 [64], and Tiny YOLO [64] to screen baggage threats contained within the scans of a publicly available GDXray dataset [72]. Gaus et al [10] utilized RetinaNet [13], Faster R-CNN [12], Mask R-CNN [11] (driven through ResNets [73], VGG-16 [74],…”

Section: Supervised Methodsmentioning

confidence: 99%

Tensor Pooling Driven Instance Segmentation Framework for Baggage Threat Recognition

Hassan¹,

Akçay²,

Bennamoun³

et al. 2021

Preprint

View full text Add to dashboard Cite

Automated systems designed for screening contraband items from the X-ray imagery are still facing difficulties with high clutter, concealment, and extreme occlusion. In this paper, we addressed this challenge using a novel multi-scale contour instance segmentation framework that effectively identifies the cluttered contraband data within the baggage X-ray scans. Unlike standard models that employ region-based or keypoint-based techniques to generate multiple boxes around objects, we propose to derive proposals according to the hierarchy of the regions defined by the contours. The proposed framework is rigorously validated on three public datasets, dubbed GDXray, SIXray, and OPIXray, where it outperforms the state-of-the-art methods by achieving the mean average precision score of 0.9779, 0.9614, and 0.8396, respectively. Furthermore, to the best of our knowledge, this is the first contour instance segmentation framework that leverages multi-scale information to

show abstract

An evaluation of deep learning based object detection strategies for threat object detection in baggage security imagery

Cited by 79 publications

References 5 publications

Automated object detection of mechanical fasteners using faster region based convolutional neural networks

Automated object detection of mechanical fasteners using faster region based convolutional neural networks

Temporal Fusion Based Mutli-scale Semantic Segmentation for Detecting Concealed Baggage Threats

Tensor Pooling Driven Instance Segmentation Framework for Baggage Threat Recognition

Contact Info

Product

Resources

About