RMOPP: Robust Multi-Objective Post-Processing for Effective Object Detection. Savargaonkar, M., Chehade, A., & Rawashdeh, S. 2, 2021. ![RMOPP: Robust Multi-Objective Post-Processing for Effective Object Detection [link]](https://bibbase.org/img/filetypes/link.svg "link")
Website abstract bibtex Over the last few decades, many architectures have been developed that harness the power of neural networks to detect objects in near real-time. Training such systems requires substantial time across multiple GPUs and massive labeled training datasets. Although the goal of these systems is generalizability, they are often impractical in real-life applications due to flexibility, robustness, or speed issues. This paper proposes RMOPP: A robust multi-objective post-processing algorithm to boost the performance of fast pre-trained object detectors with a negligible impact on their speed. Specifically, RMOPP is a statistically driven, post-processing algorithm that allows for simultaneous optimization of precision and recall. A unique feature of RMOPP is the Pareto frontier that identifies dominant possible post-processed detectors to optimize for both precision and recall. RMOPP explores the full potential of a pre-trained object detector and is deployable for near real-time predictions. We also provide a compelling test case on YOLOv2 using the MS-COCO dataset.
@article{
title = {RMOPP: Robust Multi-Objective Post-Processing for Effective Object Detection},
type = {article},
year = {2021},
websites = {http://arxiv.org/abs/2102.04582},
month = {2},
day = {8},
id = {8e40ce1f-a506-3fd3-9261-2f5142979293},
created = {2021-11-08T20:03:10.202Z},
file_attached = {false},
profile_id = {ceb22f87-8e17-3a73-8ac5-df584316b424},
last_modified = {2021-11-08T20:03:10.202Z},
read = {false},
starred = {false},
authored = {true},
confirmed = {true},
hidden = {false},
private_publication = {false},
abstract = {Over the last few decades, many architectures have been developed that harness the power of neural networks to detect objects in near real-time. Training such systems requires substantial time across multiple GPUs and massive labeled training datasets. Although the goal of these systems is generalizability, they are often impractical in real-life applications due to flexibility, robustness, or speed issues. This paper proposes RMOPP: A robust multi-objective post-processing algorithm to boost the performance of fast pre-trained object detectors with a negligible impact on their speed. Specifically, RMOPP is a statistically driven, post-processing algorithm that allows for simultaneous optimization of precision and recall. A unique feature of RMOPP is the Pareto frontier that identifies dominant possible post-processed detectors to optimize for both precision and recall. RMOPP explores the full potential of a pre-trained object detector and is deployable for near real-time predictions. We also provide a compelling test case on YOLOv2 using the MS-COCO dataset.},
bibtype = {article},
author = {Savargaonkar, Mayuresh and Chehade, Abdallah and Rawashdeh, Samir}
}
Downloads: 0
{"_id":"NR2BJzaYdZ2ySocoB","bibbaseid":"savargaonkar-chehade-rawashdeh-rmopprobustmultiobjectivepostprocessingforeffectiveobjectdetection-2021","author_short":["Savargaonkar, M.","Chehade, A.","Rawashdeh, S."],"bibdata":{"title":"RMOPP: Robust Multi-Objective Post-Processing for Effective Object Detection","type":"article","year":"2021","websites":"http://arxiv.org/abs/2102.04582","month":"2","day":"8","id":"8e40ce1f-a506-3fd3-9261-2f5142979293","created":"2021-11-08T20:03:10.202Z","file_attached":false,"profile_id":"ceb22f87-8e17-3a73-8ac5-df584316b424","last_modified":"2021-11-08T20:03:10.202Z","read":false,"starred":false,"authored":"true","confirmed":"true","hidden":false,"private_publication":false,"abstract":"Over the last few decades, many architectures have been developed that harness the power of neural networks to detect objects in near real-time. Training such systems requires substantial time across multiple GPUs and massive labeled training datasets. Although the goal of these systems is generalizability, they are often impractical in real-life applications due to flexibility, robustness, or speed issues. This paper proposes RMOPP: A robust multi-objective post-processing algorithm to boost the performance of fast pre-trained object detectors with a negligible impact on their speed. Specifically, RMOPP is a statistically driven, post-processing algorithm that allows for simultaneous optimization of precision and recall. A unique feature of RMOPP is the Pareto frontier that identifies dominant possible post-processed detectors to optimize for both precision and recall. RMOPP explores the full potential of a pre-trained object detector and is deployable for near real-time predictions. We also provide a compelling test case on YOLOv2 using the MS-COCO dataset.","bibtype":"article","author":"Savargaonkar, Mayuresh and Chehade, Abdallah and Rawashdeh, Samir","bibtex":"@article{\n title = {RMOPP: Robust Multi-Objective Post-Processing for Effective Object Detection},\n type = {article},\n year = {2021},\n websites = {http://arxiv.org/abs/2102.04582},\n month = {2},\n day = {8},\n id = {8e40ce1f-a506-3fd3-9261-2f5142979293},\n created = {2021-11-08T20:03:10.202Z},\n file_attached = {false},\n profile_id = {ceb22f87-8e17-3a73-8ac5-df584316b424},\n last_modified = {2021-11-08T20:03:10.202Z},\n read = {false},\n starred = {false},\n authored = {true},\n confirmed = {true},\n hidden = {false},\n private_publication = {false},\n abstract = {Over the last few decades, many architectures have been developed that harness the power of neural networks to detect objects in near real-time. Training such systems requires substantial time across multiple GPUs and massive labeled training datasets. Although the goal of these systems is generalizability, they are often impractical in real-life applications due to flexibility, robustness, or speed issues. This paper proposes RMOPP: A robust multi-objective post-processing algorithm to boost the performance of fast pre-trained object detectors with a negligible impact on their speed. Specifically, RMOPP is a statistically driven, post-processing algorithm that allows for simultaneous optimization of precision and recall. A unique feature of RMOPP is the Pareto frontier that identifies dominant possible post-processed detectors to optimize for both precision and recall. RMOPP explores the full potential of a pre-trained object detector and is deployable for near real-time predictions. We also provide a compelling test case on YOLOv2 using the MS-COCO dataset.},\n bibtype = {article},\n author = {Savargaonkar, Mayuresh and Chehade, Abdallah and Rawashdeh, Samir}\n}","author_short":["Savargaonkar, M.","Chehade, A.","Rawashdeh, S."],"urls":{"Website":"http://arxiv.org/abs/2102.04582"},"biburl":"https://bibbase.org/service/mendeley/ceb22f87-8e17-3a73-8ac5-df584316b424","bibbaseid":"savargaonkar-chehade-rawashdeh-rmopprobustmultiobjectivepostprocessingforeffectiveobjectdetection-2021","role":"author","metadata":{"authorlinks":{}},"downloads":0},"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/ceb22f87-8e17-3a73-8ac5-df584316b424","dataSources":["2252seNhipfTmjEBQ"],"keywords":[],"search_terms":["rmopp","robust","multi","objective","post","processing","effective","object","detection","savargaonkar","chehade","rawashdeh"],"title":"RMOPP: Robust Multi-Objective Post-Processing for Effective Object Detection","year":2021}