Top View Person Detection and Counting for Low Compute Embedded Platforms

Maheshwari, Prashant; Alex, Doney; Banerjee, Somnath; Behera, Saurav; Panda, Satyajit

doi:10.1145/3301506.3301548

Cited by 3 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The comparative analysis is summarised in Table 1 to highlight representative published results related to top-view human detection, in comparison with the proposed solution in this paper (highlighted in green). In one study, Maheshwari et al [28] have used state-ofthe-art Ada-Boost classifier for detection and kalman filter for tracking of people from top-view perspective. Although a high accuracy of 97% was achieved, and the authors mentioned different altitudes, the results are not demonstrated at high altitudes and cluttered environments, which make the detection significantly more complex.…”

Section: Previous Workmentioning

confidence: 99%

See 1 more Smart Citation

Machine-learning-based top-view safety monitoring of ground workforce on complex industrial sites

Golkarnarenji

Martinez‐Alpiste

Wang

et al. 2021

Neural Comput & Applic

View full text Add to dashboard Cite

Telescopic cranes are powerful lifting facilities employed in construction, transportation, manufacturing and other industries. Since the ground workforce cannot be aware of their surrounding environment during the current crane operations in busy and complex sites, accidents and even fatalities are not avoidable. Hence, deploying an automatic and accurate top-view human detection solution would make significant improvements to the health and safety of the workforce on such industrial operational sites. The proposed method (CraneNet) is a new machine learning empowered solution to increase the visibility of a crane operator in complex industrial operational environments while addressing the challenges of human detection from top-view on a resource-constrained small-form PC to meet the space constraint in the operator’s cabin. CraneNet consists of 4 modified ResBlock-D modules to fulfill the real-time requirements. To increase the accuracy of small humans at high altitudes which is crucial for this use-case, a PAN (Path Aggregation Network) was designed and added to the architecture. This enhances the structure of CraneNet by adding a bottom-up path to spread the low-level information. Furthermore, three output layers were employed in CraneNet to further improve the accuracy of small objects. Spatial Pyramid Pooling (SPP) was integrated at the end of the backbone stage which increases the receptive field of the backbone, thereby increasing the accuracy. The CraneNet has achieved 92.59% of accuracy at 19 FPS on a portable device. The proposed machine learning model has been trained with the Standford Drone Dataset and Visdrone 2019 to further show the efficacy of the smart crane approach. Consequently, the proposed system is able to detect people in complex industrial operational areas from a distance up to 50 meters between the camera and the person. This system is also applicable to the detection of any other objects from an overhead camera.

show abstract

Section: Previous Workmentioning

confidence: 99%

“…However, the result has not been reported at high altitudes which makes the detection more challenging. All of the studies above except [28] have been conducted on powerful PCs and not suitable on embedded systems such as Jetson Xavier.…”

Section: Previous Workmentioning

confidence: 99%

Machine-learning-based top-view safety monitoring of ground workforce on complex industrial sites

Golkarnarenji

Martinez‐Alpiste

Wang

et al. 2021

Neural Comput & Applic

View full text Add to dashboard Cite

show abstract

“…[9] has stated that the explicit visibility of cameras or surveillance measures makes employees seem worried in the workplace leading to reduced creativity. To address this issue, in our research we seek to reduce the risk of documenting individuals' actions by installing an privacypreserving surveillance system at the room ceiling similar to [11]. Additionally, since the camera has a physical lens cover it may look less intimidating to the employees.…”

Section: Related Workmentioning

confidence: 99%

“…In a regional qualitative study carried out by [24], it was noted that individuals are not aware of whether there are any rules to demand clarification on the usage of the gathered data from the observers. Apart from awareness of data usage among users, [11], [12] demonstrated how fragile an individuals' privacy is under state-of-the-art technology-enabled video surveillance. [6] states that any kind of human skin (face, skin tone) under the surveillance footage can behave as a feature that can lead to revealing the provenance of both, the video and people in the footage.…”

Section: Related Workmentioning

confidence: 99%

Explicitly Privacy-Aware Space Usage Analysis

Jha

Mayer

Schneider

2020

2020 IEEE 19th International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)

View full text Add to dashboard Cite

Surveillance in private and public spaces provides observers with information that can enhance protection and efficiency but usually infringes upon the privacy of the individuals and groups. These informational privacy risks are centered on users' perceived and design-induced threats. They cannot be removed completely but can be minimized using suitable anonymization techniques. To minimize the users' informational privacy threats, we designed a privacy-aware surveillance system that gives the users leverage over the anonymization filters, to physically adjust the opaqueness of the camera lens used in the prototype according to their privacy requirements. We implement our prototype in the context of office space surveillance, where the proposed solution considers privacy requirements in such environments to improve users' trust in the surveillance system and reduce their privacy concerns.

show abstract