Abstract:In recent years, intelligent driving navigation and security monitoring have made considerable progress with the help of deep Convolutional Neural Networks (CNNs). As one of the state-of-the-art perception approaches, semantic segmentation unifies distinct detection tasks widely desired by both autonomous driving and security monitoring. Currently, semantic segmentation shows remarkable efficiency and reliability in standard scenarios such as daytime scenes with favorable illumination conditions. However, in face of adverse conditions such as the nighttime, semantic segmentation loses its accuracy significantly. One of the main causes of the problem is the lack of sufficient annotated segmentation datasets of nighttime scenes. In this paper, we propose a framework to alleviate the accuracy decline when semantic segmentation is taken to adverse conditions by using Generative Adversarial Networks (GANs). To bridge the daytime and nighttime image domains, we made key observation that compared to datasets in adverse conditions, there are considerable amount of segmentation datasets in standard conditions such as BDD and our collected ZJU datasets. Our GAN-based nighttime semantic segmentation framework includes two methods. In the first method, GANs were used to translate nighttime images to the daytime, thus semantic segmentation can be performed using robust models already trained on daytime datasets. In another method, we use GANs to translate different ratio of daytime images in the dataset to the nighttime but still with their labels. In this sense, synthetic nighttime segmentation datasets can be generated to yield models prepared to operate at nighttime conditions robustly. In our experiment, the later method significantly boosts the performance at the nighttime evidenced by quantitative results using Intersection over Union (IoU) and Pixel Accuracy (Acc). We show that the performance varies with respect to the proportion of synthetic nighttime images in the dataset, where the sweet spot corresponds to most robust performance across the day and night. The proposed framework not only makes contribution to the optimization of visual perception in intelligent vehicles, but also can be applied to diverse navigational assistance systems.

Energy-efficient real-time visual image adversarial generation and processing algorithm for new energy vehicles

MND-GAN: A Research on Image Deblurring Algorithm Based on Generative Adversarial Network

Attack Deterministic Conditional Image Generative Models for Diverse and Controllable Generation

Towards Understanding the Generative Capability of Adversarially Robust Classifiers

Virtual View Generation Based on 3D-Dense-attentive GAN Networks

See Clearer at Night: Towards Robust Nighttime Semantic Segmentation Through Day-Night Image Conversion

Joint Generative Image Deblurring Aided by Edge Attention Prior and Dynamic Kernel Selection

Enhanced Attentive Generative Adversarial Network for Single-Image Deraining.

Using Vehicle Synthesis Generative Adversarial Networks to Improve Vehicle Detection in Remote Sensing Images

Perceptual Enhancement for Autonomous Vehicles: Restoring Visually Degraded Images for Context Prediction via Adversarial Training

PhysGAN: Generating Physical-World-Resilient Adversarial Examples for Autonomous Driving

All-Day Vehicle Detection From Surveillance Videos Based on Illumination-Adjustable Generative Adversarial Network

Rendering the Image of Glare Effect Based on Paired and Unpaired Dual Generative Adversarial Network

Learning Image Attacks toward Vision Guided Autonomous Vehicles

SST-GAN: Single Sample-based Realistic Traffic Image Generation for Parallel Vision

DeepRoad: GAN-based Metamorphic Autonomous Driving System Testing

Domain‐adapted Driving Scene Understanding with Uncertainty‐aware and Diversified Generative Adversarial Networks

EVGen: Adversarial Networks for Learning Electric Vehicle Charging Loads and Hidden Representations

GAN-Based Day-to-Night Image Style Transfer for Nighttime Vehicle Detection

Generating Traffic Scene with Deep Convolutional Generative Adversarial Networks

ADGAN: Protect Your Location Privacy in Camera Data of Auto-Driving Vehicles