Vision-Based Semantic Segmentation in Scene Understanding for Autonomous Driving: Recent Achievements, Challenges, and Outlooks K. Muhammad, T. Hussain, H. Ullah, J. D. Ser, M. Rezaei, N. Kumar, M. Hijji, P. Bellavista, and V. H. C. de Albuquerque Scene understanding in autonomous driving utilizes sensory data for contextual information extraction and decision making. Related research directions include person/vehicle detection/segmentation, their transition analysis, lane change, and turns detection, among others. Unfortunately, these tasks seem insufficient to completely develop fully autonomous vehicles, i.e., achieving level-5 autonomy, traveling just like human-controlled cars. This statement is among the conclusions drawn from this review paper: scene understanding for autonomous driving cars using vision sensors still requires significant improvements. Specifically, the authors define, analyze, and examine current achievements in scene understanding, covering the generic scene understanding pipeline, investigating state-of-the-art performance levels, informing about the time complexity analysis of avant-garde models, and highlighting major triumphs and limitations identified by recent research efforts. The authors also include a thorough discussion on available datasets and challenges that, even if lately confronted by researchers, remain open to date. Finally, research directions are offered to welcome researchers/practitioners to this exciting domain.
Scanning the Issue
IEEE Transactions on Intelligent Transportation Systems ; 23 , 12 ; 22660-22693
2022-12-01
395941 byte
Article (Journal)
Electronic Resource
English
IEEE | 2021
|IEEE | 2017
|IEEE | 2017
|IEEE | 2017
|IEEE | 2016
|