Deformable Convolution Based Road Scene Semantic Segmentation of Fisheye Images in Autonomous Driving

TL;DR

This paper demonstrates that Deformable Convolutional Neural Networks significantly improve semantic segmentation accuracy of fisheye images in autonomous driving by effectively handling geometric distortions and complex spatial relationships.

Contribution

It introduces the application of Deformable CNNs to fisheye image segmentation, showing superior performance over traditional CNN architectures in autonomous driving scenarios.

Findings

Deformable CNNs outperform traditional CNNs in mIoU scores.

Incorporating Deformable convolutions improves segmentation of distorted fisheye images.

Different loss functions help address class imbalance in segmentation tasks.

Abstract

This study investigates the effectiveness of modern Deformable Convolutional Neural Networks (DCNNs) for semantic segmentation tasks, particularly in autonomous driving scenarios with fisheye images. These images, providing a wide field of view, pose unique challenges for extracting spatial and geometric information due to dynamic changes in object attributes. Our experiments focus on segmenting the WoodScape fisheye image dataset into ten distinct classes, assessing the Deformable Networks' ability to capture intricate spatial relationships and improve segmentation accuracy. Additionally, we explore different loss functions to address class imbalance issues and compare the performance of conventional CNN architectures with Deformable Convolution-based CNNs, including Vanilla U-Net and Residual U-Net architectures. The significant improvement in mIoU score resulting from integrating Deformable CNNs demonstrates their effectiveness in handling the geometric distortions present in fisheye imagery, exceeding the performance of traditional CNN architectures. This underscores the significant role of Deformable convolution in enhancing semantic segmentation performance for fisheye imagery.