Survey of real-time object detection networks including versatility, robustness, resources, energy
Survey of real-time object detection networks including versatility, robustness, resources, energy
A Comprehensive Study of Real-Time Object Detection Networks Across Multiple Domains: A Survey
arXiv paper abstract https://arxiv.org/abs/2208.10895
arXiv PDF paper https://arxiv.org/pdf/2208.10895.pdf
Deep neural network based object detectors are continuously evolving and are used in a multitude of applications, each having its own set of requirements.
… Real-time detectors, which are a necessity in high-impact real-world applications, are continuously proposed, but they overemphasize the improvements in accuracy and speed while other capabilities such as versatility, robustness, resource and energy efficiency are omitted.
… conduct a comprehensive study on multiple real-time detectors (anchor-, keypoint-, and transformer-based) on a wide range of datasets and report results on an extensive set of metrics.
… also study the impact of variables such as image size, anchor dimensions, confidence thresholds, and architecture layers on the overall performance.
… analyze the robustness … against distribution shifts, natural corruptions, and adversarial attacks … provide a calibration analysis to gauge the reliability of the predictions.
… To … gauge the capability … in critical real-time applications … report the performance after deploying … on edge devices …
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
