A Pipeline-Based Approach for Object Detection on Resource Constrained Internet of Things Devices
Primary Area: optimization
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Keywords: Edge Computing, Artificial Intelligence, Internet of Things, Computer Vision
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
TL;DR: A pipeline-based approach for object detection on resource constrained internet of things devices for cognitive edge computing applications.
Abstract: Object detection with computer vision and convolutional neural networks on resource constrained devices can be challenging. The limited power and processing capacity of these devices complicates the use of deep neural networks and other object detection methods. To address this problem, we propose a pipeline-based approach. We introduce a multi-step detection pipeline considering the size of the objects to be detected and the correlation among them. To evaluate the performance of this approach, we test it in a collaborative smart surveillance system employing edge computing and the internet of things paradigm. Additionally, field testing was conducted considering real world surveillance scenarios. Results showed that the introduction of the pipeline-based processing improved the execution time by a factor of 3 and produced a significant improvement on the mean average precision.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 5762
Loading