FSNet: A Failure Detection Framework for Semantic Segmentation

  title={FSNet: A Failure Detection Framework for Semantic Segmentation},
  author={Q. Rahman and Niko Sunderhauf and Peter Corke and Feras Dayoub},
  journal={IEEE Robotics and Automation Letters},
Semantic segmentation is an important task that helps autonomous vehicles understand their surroundings and navigate safely. However, during deployment, even the most mature segmentation models are vulnerable to various external factors that can degrade the segmentation performance with potentially catastrophic consequences for the vehicle and its surroundings. To address this issue, we propose a failure detection framework to identify pixel-level misclassification. We do so by exploiting… 

Figures and Tables from this paper

See Yourself in Others: Attending Multiple Tasks for Own Failure Detection

This work proposes an attention-based failure detection approach by exploiting the correlations among multiple tasks to infers task failures by evaluating the individual prediction, across multiple visual perception tasks for different regions in an image.

Energy Efficient Automatic Streetlight Controlling System using Semantic Segmentation

This study aims to develop a novel streetlight management system powered by computer vision technology mounted with the CCTV camera that allows the LED streetlight to automatically light up with proper brightness by recognizing the presence of pedestrians or vehicles and reversely dimming the streetlight in their absence by semantic image segmentation from video.

Failure Detection for Motion Prediction of Autonomous Driving: An Uncertainty Perspective

 Abstract — Motion prediction is essential for safe and efficient autonomous driving. However, the inexplicability and uncertainty of complex artificial intelligence models may lead to unpredictable

Monitoring of Perception Systems: Deterministic, Probabilistic, and Learning-based Fault Detection and Identification

This paper formalizes the problem of runtime fault detection and identification in perception systems and presents a framework to model diagnostic information using a diagnostic graph, and provides a set of deterministic, probabilistic, and learning-based algorithms that use diagnostic graphs to perform fault detection.

Intelligence of Autonomous Vehicles: A Concise Revisit

A concise review of the state-of-the-art techniques to improve the performance of autonomous vehicles is presented.



Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation

This paper systematically study failure and anomaly detection for semantic segmentation and proposes a unified framework, consisting of two modules, to address these two related problems.

Introspective Failure Prediction for Semantic Image Segmentation

In a precision-recall analysis, the proposed method outperforms two state-of-the-art uncertainty estimation methods by 3.2% and 6.7% while requiring significantly less resources during inference.

Survey on semantic segmentation using deep learning techniques

Deep Multi-Modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challenges

This review paper attempts to systematically summarize methodologies and discuss challenges for deep multi-modal object detection and semantic segmentation in autonomous driving with an overview of on-board sensors on test vehicles, open datasets, and background information.

Speeding up Semantic Segmentation for Autonomous Driving

A novel deep network architecture for image segmentation that keeps the high accuracy while being efficient enough for embedded devices is proposed, and achieves higher segmentation accuracy than other networks that are tailored to embedded devices.

A Brief Survey on Semantic Segmentation with Deep Learning

Road segmentation for all-day outdoor robot navigation

Convolutional CRFs for Semantic Segmentation

This work proposes to add the assumption of conditional independence to the framework of fully-connected CRFs, which allows the inference to be reformulated in terms of convolutions, which can be implemented highly efficiently on GPUs.

QualityNet: Segmentation quality evaluation with deep convolutional networks

A deep convolutional network is proposed for quality evaluation of object segmentation and the weighted mask layer is proposed to utilize both the information of foreground and background to help obtain better segmentation result based on the predicted segmentation quality score.