Deep semantic segmentation for automated driving: Taxonomy, roadmap and challenges

By Mennatullah Siam, Sara Elkerdawy, Martin Jägersand, Senthil Kumar Yogamani. 2017

181 citations recorded by the source corpus.

Semantic segmentation was seen as a challenging computer vision problem few years ago. Due to recent advancements in deep learning, relatively accurate solutions are now possible for its use in automated driving. In this paper, the semantic segmentation problem is explored from the perspective of automated driving. Most of the current semantic segmentation algorithms are designed for generic images and do not incorporate prior structure and end goal for automated driving. First, the paper begins with a generic taxonomic survey of semantic segmentation algorithms and then discusses how it fits in the context of automated driving. Second, the particular challenges of deploying it into a safety system which needs high level of accuracy and robustness are listed. Third, different alternatives instead of using an independent semantic segmentation module are explored. Finally, an empirical evaluation of various semantic segmentation architectures was performed on CamVid dataset in terms of accuracy and speed. This paper is a preliminary shorter version of a more detailed survey which is work in progress.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

WoodScape: A Multi-Task, Multi-Camera Fisheye Dataset for Autonomous Driving (2019, neutral): “Semantic segmentation networks for autonomous driving [47] have been successfully trained directly on fisheye images in [12, 45].”
Deep Reinforcement Learning for Autonomous Driving: A Survey (2020, background): “This is achieved by a combination of several perception tasks like semantic segmentation [3], [4], motion estimation [5], depth estimation [6], soiling detection [7], etc which can be efﬁciently uniﬁed into a multi-task model [8], [9].”
Segmenter: Transformer for Semantic Segmentation (2021, background): “Semantic segmentation is a challenging computer vision problem with a wide range of applications including autonomous driving, robotics, augmented reality, image editing, medical imaging and many others [26, 27, 43].”
NeurAll: Towards a Unified Model for Visual Perception in Automated Driving (2019, background): “[5] provided a detailed overview of semantic segmentation for automated driving applications.”
Machine Learning and Deep Learning frameworks and libraries for large-scale data mining: a survey (2019, background): “…et al. 2017) or galaxies (Dieleman et al. 2015)], image segmentation [with applications to crowd counting (Sam et al. 2017), autonomous driving (Siam et al. 2017) or satellite imagery (Mnih 2013)], along with many other miscellaneous applications Table 1 DL timeline through the most well-known…”
Yes, we GAN: Applying Adversarial Techniques for Autonomous Driving (2019, background): “Progress in Deep learning has accelerated the maturity of autonomous driving systems [1].”
Knowledge Distillation for Semantic Segmentation Using Channel and Spatial Correlations and Adaptive Cross Entropy (2020, methodology): “[37] adopted the fire module proposed in [38] and parallel dilated convolution.”
RGB and LiDAR fusion based 3D Semantic Segmentation for Autonomous Driving (2019, methodology): “A detailed survey for image semantic segmentation for autonomous driving is presented in [17] and efﬁcient design techniques were discussed in [2].”
NeurAll: Towards a Unified Visual Perception Model for Automated Driving (2019, background): “[5] provided a detailed overview of semantic segmentation for automated driving applications.”
Asymmetric Non-Local Neural Networks for Semantic Segmentation (2019, background): “A encoder generally reduces the spatial size of feature maps to enlarge the receptive ﬁeld.”
SDDNet: Real-Time Crack Segmentation (2020, background): “Object segmentation has been successfully implemented in medical image analysis [17], self-driving vehicles [18], and so on.”
AuxNet: Auxiliary tasks enhanced Semantic Segmentation for Automated Driving (2019, background): “A detailed survey of semantic segmentation for automated driving is presented in (Siam et al., 2017).”