An Efficient Intersection Over Union Algorithm for 3D Object Detection

Quite recently, the combination of improved accuracy, flexibility, available datasets, state-of-the-art architectures, and diverse applications has contributed to the widespread adoption and popularity of deep learning-based object detection techniques, particularly for many applications in computer...

Full description

Saved in:
Bibliographic Details
Main Authors: Sazan Ali Kamal Mohammed, Mohd Zulhakimi Ab Razak, Abdul Hadi Abd Rahman, Maria Abu Bakar
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10750188/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846160745319890944
author Sazan Ali Kamal Mohammed
Mohd Zulhakimi Ab Razak
Abdul Hadi Abd Rahman
Maria Abu Bakar
author_facet Sazan Ali Kamal Mohammed
Mohd Zulhakimi Ab Razak
Abdul Hadi Abd Rahman
Maria Abu Bakar
author_sort Sazan Ali Kamal Mohammed
collection DOAJ
description Quite recently, the combination of improved accuracy, flexibility, available datasets, state-of-the-art architectures, and diverse applications has contributed to the widespread adoption and popularity of deep learning-based object detection techniques, particularly for many applications in computer vision and emerging applications. An important metric in this discipline is the Intersection over Union (IoU) loss function, which is common and extensively being used in the boundary analysis. The majority of enhanced loss functions often alter the penalty terms when solving the zero-gradient issue of IoU in non-overlapping situations. Therefore, in this study, we first analyze the existing IoU loss functions in order to propose an efficient intersection over union (EIoU) algorithm. The proposed EIoU is capable of providing the gradient value even though the bounded predicted box and the ground truth target box do not overlap. In this case, we propose a steady optimization procedure (SOP) for the EIoU loss to gradually approach the minimum value. In addition, we also propose an angle extension of the EIoU algorithm as a significant refinement tool. The EIoU loss function is inversely correlated with the size of the bounded boxes and responds quickly to a small object size, which favors the detection of small objects in a 3D point cloud, where fewer iterations are required to make the bounded box regress well. The proposed method records error values at least 78% better than other types of IoUs. Similarly, EIoU reaches 72.3%, 17.4%, and 85.5% higher values for intersection, separation, and inclusion location conditions, respectively, as compared to other IoU types, indicating that EIoU is sensitive to the change in boxes dimensions and hence, leads to fast convergence between the predicted and ground truth boxes. Experimental results on the KITTI datasets for 3D object detection achieves 11.5% and 12.7% improvement for SECOND+EIoU test condition as compared SECOND only and at least 8% and 6.9% better than other the state-of-the-art 3D methods for moderate car and pedestrian cases, respectively, which proved that our proposed techniques have significantly improved the precision accuracy of learning and the performance of 3D object detection.
format Article
id doaj-art-29e201dbf8f24975b81bec10956431e6
institution Kabale University
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-29e201dbf8f24975b81bec10956431e62024-11-22T00:01:08ZengIEEEIEEE Access2169-35362024-01-011216976816978610.1109/ACCESS.2024.349576110750188An Efficient Intersection Over Union Algorithm for 3D Object DetectionSazan Ali Kamal Mohammed0https://orcid.org/0000-0002-7119-1575Mohd Zulhakimi Ab Razak1https://orcid.org/0000-0001-7590-669XAbdul Hadi Abd Rahman2https://orcid.org/0000-0002-0261-073XMaria Abu Bakar3https://orcid.org/0000-0002-9745-3334Institute of Microengineering and Nanoelectronics (IMEN), Universiti Kebangsaan Malaysia, Bangi, MalaysiaInstitute of Microengineering and Nanoelectronics (IMEN), Universiti Kebangsaan Malaysia, Bangi, MalaysiaCenter for Artificial Intelligence Technology, Universiti Kebangsaan Malaysia, Bangi, MalaysiaInstitute of Microengineering and Nanoelectronics (IMEN), Universiti Kebangsaan Malaysia, Bangi, MalaysiaQuite recently, the combination of improved accuracy, flexibility, available datasets, state-of-the-art architectures, and diverse applications has contributed to the widespread adoption and popularity of deep learning-based object detection techniques, particularly for many applications in computer vision and emerging applications. An important metric in this discipline is the Intersection over Union (IoU) loss function, which is common and extensively being used in the boundary analysis. The majority of enhanced loss functions often alter the penalty terms when solving the zero-gradient issue of IoU in non-overlapping situations. Therefore, in this study, we first analyze the existing IoU loss functions in order to propose an efficient intersection over union (EIoU) algorithm. The proposed EIoU is capable of providing the gradient value even though the bounded predicted box and the ground truth target box do not overlap. In this case, we propose a steady optimization procedure (SOP) for the EIoU loss to gradually approach the minimum value. In addition, we also propose an angle extension of the EIoU algorithm as a significant refinement tool. The EIoU loss function is inversely correlated with the size of the bounded boxes and responds quickly to a small object size, which favors the detection of small objects in a 3D point cloud, where fewer iterations are required to make the bounded box regress well. The proposed method records error values at least 78% better than other types of IoUs. Similarly, EIoU reaches 72.3%, 17.4%, and 85.5% higher values for intersection, separation, and inclusion location conditions, respectively, as compared to other IoU types, indicating that EIoU is sensitive to the change in boxes dimensions and hence, leads to fast convergence between the predicted and ground truth boxes. Experimental results on the KITTI datasets for 3D object detection achieves 11.5% and 12.7% improvement for SECOND+EIoU test condition as compared SECOND only and at least 8% and 6.9% better than other the state-of-the-art 3D methods for moderate car and pedestrian cases, respectively, which proved that our proposed techniques have significantly improved the precision accuracy of learning and the performance of 3D object detection.https://ieeexplore.ieee.org/document/10750188/Average precision3D object detectionIoUloss functionpoint cloudinclusion
spellingShingle Sazan Ali Kamal Mohammed
Mohd Zulhakimi Ab Razak
Abdul Hadi Abd Rahman
Maria Abu Bakar
An Efficient Intersection Over Union Algorithm for 3D Object Detection
IEEE Access
Average precision
3D object detection
IoU
loss function
point cloud
inclusion
title An Efficient Intersection Over Union Algorithm for 3D Object Detection
title_full An Efficient Intersection Over Union Algorithm for 3D Object Detection
title_fullStr An Efficient Intersection Over Union Algorithm for 3D Object Detection
title_full_unstemmed An Efficient Intersection Over Union Algorithm for 3D Object Detection
title_short An Efficient Intersection Over Union Algorithm for 3D Object Detection
title_sort efficient intersection over union algorithm for 3d object detection
topic Average precision
3D object detection
IoU
loss function
point cloud
inclusion
url https://ieeexplore.ieee.org/document/10750188/
work_keys_str_mv AT sazanalikamalmohammed anefficientintersectionoverunionalgorithmfor3dobjectdetection
AT mohdzulhakimiabrazak anefficientintersectionoverunionalgorithmfor3dobjectdetection
AT abdulhadiabdrahman anefficientintersectionoverunionalgorithmfor3dobjectdetection
AT mariaabubakar anefficientintersectionoverunionalgorithmfor3dobjectdetection
AT sazanalikamalmohammed efficientintersectionoverunionalgorithmfor3dobjectdetection
AT mohdzulhakimiabrazak efficientintersectionoverunionalgorithmfor3dobjectdetection
AT abdulhadiabdrahman efficientintersectionoverunionalgorithmfor3dobjectdetection
AT mariaabubakar efficientintersectionoverunionalgorithmfor3dobjectdetection