APSIPA Transactions on Signal and Information Processing > Vol 12 > Issue 1

A Survey of Efficient Deep Learning Models for Moving Object Segmentation

Bingxin Hou, Department of Computer Science and Engineering, Santa Clara University, USA, Ying Liu, Department of Computer Science and Engineering, Santa Clara University, USA, Nam Ling, Department of Computer Science and Engineering, Santa Clara University, USA, nling@scu.edu , Yongxiong Ren, Kwai, Inc., USA, Lingzhi Liu, Kwai, Inc., USA
 
Suggested Citation
Bingxin Hou, Ying Liu, Nam Ling, Yongxiong Ren and Lingzhi Liu (2023), "A Survey of Efficient Deep Learning Models for Moving Object Segmentation", APSIPA Transactions on Signal and Information Processing: Vol. 12: No. 1, e2. http://dx.doi.org/10.1561/116.00000140

Publication Date: 03 Jan 2023
© 2023 B. Hou, Y. Liu, N. Ling, Y. Ren and L. Liu
 
Subjects
 
Keywords
Background subtractionchange detectiondeep learningefficient model designlightweight modelmoving object segmentationreal-time processingvideo object segmentation
 

Share

Open Access

This is published under the terms of CC BY-NC.

Downloaded: 2157 times

In this article:
Introduction 
Existing MOS Approaches 
Efficient MOS Model Design 
Datasets 
Metrics 
Performance Comparison of Efficient MOS Models 
Challenges and Future Directions 
Conclusion 
References 

Abstract

Moving object segmentation (MOS) is the process of identifying dynamic objects from video frames, such as moving vehicles or pedestrians, while discarding the background. It plays an essential role in many real-world applications such as autonomous driving, mobile robots, and surveillance systems. With the availability of a huge amount of data and the development of powerful computing infrastructure, deep learning-based methods have shown remarkable improvements in MOS tasks. However, as the dimension of data becomes higher and the network architecture becomes more complicated, deep learning-based MOS models are computationally intensive, which limits their deployment on resource-constrained devices and in delay-sensitive applications. Therefore, more research started to develop fast and lightweight models. This paper aims to provide a comprehensive review of deep learning-based MOS models, with a focus on efficient model design techniques. We summarize a variety of MOS datasets, and conduct a thorough review of segmentation accuracy metrics and model efficiency metrics. Most importantly, we compare the performance of efficient MOS models on popular datasets, identify competitive models and analyze their essential techniques. Finally, we point out existing challenges and present future research directions.

DOI:10.1561/116.00000140