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.

Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation

Troubleshooting image segmentation models with human-in-the-loop

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

Anomaly-Aware Semantic Segmentation by Leveraging Synthetic-Unknown Data

Diagnostics in Semantic Segmentation

A Novel Method for Road Anomaly Objects Detection in the Traffic Environment With Multi-Mechanism Fusion

Reliability in Semantic Segmentation: Can We Use Synthetic Data?

Anomaly Discovery in Semantic Segmentation via Distillation Comparison Networks

Unsupervised Anomaly Segmentation using Image-Semantic Cycle Translation

Synth4Seg -- Learning Defect Data Synthesis for Defect Segmentation using Bi-level Optimization

Automated Evaluation of Semantic Segmentation Robustness for Autonomous Driving

Automatic Discovery and Assessment of Interpretable Systematic Errors in Semantic Segmentation

AnomalyControl: Learning Cross-modal Semantic Features for Controllable Anomaly Synthesis

Synthetic Convolutional Features for Improved Semantic Segmentation

Improving Synthetic to Realistic Semantic Segmentation with Parallel Generative Ensembles for Autonomous Urban Driving

Anomaly-Aware Semantic Segmentation via Style-Aligned OoD Augmentation

Comparative Benchmarking of Failure Detection Methods in Medical Image Segmentation: Unveiling the Role of Confidence Aggregation

Trusting Semantic Segmentation Networks

Threshold-adaptive Unsupervised Focal Loss for Domain Adaptation of Semantic Segmentation

ErrorAug: Making Errors to Find Errors in Semantic Segmentation

SegmentMeIfYouCan: A Benchmark for Anomaly Segmentation