Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle
Object detection is an important aspect for autonomous driving vehicles (ADV), which may comprise of a machine learning model that detects a range of classes. As the deployment of ADV widens globally, the variety of objects to be detected may increase beyond the designated range of classes. Continua...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-11-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/20/23/6777 |
_version_ | 1797546555220164608 |
---|---|
author | Jeng-Lun Shieh Qazi Mazhar ul Haq Muhamad Amirul Haq Said Karam Peter Chondro De-Qin Gao Shanq-Jang Ruan |
author_facet | Jeng-Lun Shieh Qazi Mazhar ul Haq Muhamad Amirul Haq Said Karam Peter Chondro De-Qin Gao Shanq-Jang Ruan |
author_sort | Jeng-Lun Shieh |
collection | DOAJ |
description | Object detection is an important aspect for autonomous driving vehicles (ADV), which may comprise of a machine learning model that detects a range of classes. As the deployment of ADV widens globally, the variety of objects to be detected may increase beyond the designated range of classes. Continual learning for object detection essentially ensure a robust adaptation of a model to detect additional classes on the fly. This study proposes a novel continual learning method for object detection that learns new object class(es) along with cumulative memory of classes from prior learning rounds to avoid any catastrophic forgetting. The results of PASCAL VOC 2007 have suggested that the proposed ER method obtains 4.3% of mAP drop compared against the all-classes learning, which is the lowest amongst other prior arts. |
first_indexed | 2024-03-10T14:31:28Z |
format | Article |
id | doaj.art-dc8ebc54cda747738721eaecb43e1970 |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-10T14:31:28Z |
publishDate | 2020-11-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-dc8ebc54cda747738721eaecb43e19702023-11-20T22:32:49ZengMDPI AGSensors1424-82202020-11-012023677710.3390/s20236777Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving VehicleJeng-Lun Shieh0Qazi Mazhar ul Haq1Muhamad Amirul Haq2Said Karam3Peter Chondro4De-Qin Gao5Shanq-Jang Ruan6Department of Electronic and Computer Engineering, National Taiwan University of Science and Technology, Taipei 106, TaiwanDepartment of Electronic and Computer Engineering, National Taiwan University of Science and Technology, Taipei 106, TaiwanDepartment of Electronic and Computer Engineering, National Taiwan University of Science and Technology, Taipei 106, TaiwanDepartment of Electronic and Computer Engineering, National Taiwan University of Science and Technology, Taipei 106, TaiwanInformation and Communications Research Laboratories, Embedded Vision and Graphics Technology Department, Division for Embedded System and SoC Technology, Industrial Technology Research Institute, Hsinchu 31057, TaiwanInformation and Communications Research Laboratories, Embedded Vision and Graphics Technology Department, Division for Embedded System and SoC Technology, Industrial Technology Research Institute, Hsinchu 31057, TaiwanDepartment of Electronic and Computer Engineering, National Taiwan University of Science and Technology, Taipei 106, TaiwanObject detection is an important aspect for autonomous driving vehicles (ADV), which may comprise of a machine learning model that detects a range of classes. As the deployment of ADV widens globally, the variety of objects to be detected may increase beyond the designated range of classes. Continual learning for object detection essentially ensure a robust adaptation of a model to detect additional classes on the fly. This study proposes a novel continual learning method for object detection that learns new object class(es) along with cumulative memory of classes from prior learning rounds to avoid any catastrophic forgetting. The results of PASCAL VOC 2007 have suggested that the proposed ER method obtains 4.3% of mAP drop compared against the all-classes learning, which is the lowest amongst other prior arts.https://www.mdpi.com/1424-8220/20/23/6777continual learningone-stage object detectionautonomous driving vehicles |
spellingShingle | Jeng-Lun Shieh Qazi Mazhar ul Haq Muhamad Amirul Haq Said Karam Peter Chondro De-Qin Gao Shanq-Jang Ruan Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle Sensors continual learning one-stage object detection autonomous driving vehicles |
title | Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle |
title_full | Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle |
title_fullStr | Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle |
title_full_unstemmed | Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle |
title_short | Continual Learning Strategy in One-Stage Object Detection Framework Based on Experience Replay for Autonomous Driving Vehicle |
title_sort | continual learning strategy in one stage object detection framework based on experience replay for autonomous driving vehicle |
topic | continual learning one-stage object detection autonomous driving vehicles |
url | https://www.mdpi.com/1424-8220/20/23/6777 |
work_keys_str_mv | AT jenglunshieh continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle AT qazimazharulhaq continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle AT muhamadamirulhaq continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle AT saidkaram continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle AT peterchondro continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle AT deqingao continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle AT shanqjangruan continuallearningstrategyinonestageobjectdetectionframeworkbasedonexperiencereplayforautonomousdrivingvehicle |