# DriveGuard: Robustification of Automated Driving Systems with Deep Spatio-Temporal Convolutional Autoencoder

arXiv: https://arxiv.org/abs/2111.03480 · HTML: https://mlsanity.com/paper/2111.03480/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 45 references that are in the map.
- First posted on arXiv 5 Nov 2021.

## Builds on (most cited first, 19 of 19)

- [The Cityscapes Dataset for Semantic Urban Scene Understanding](https://mlsanity.com/paper/1604.01685/) (2016)
- [Rethinking Atrous Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1706.05587/) (2017)
- [Image Super-Resolution Using Deep Convolutional Networks](https://mlsanity.com/paper/1501.00092/) (2014)
- [Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey](https://mlsanity.com/paper/1801.00553/) (2018)
- [Toward Convolutional Blind Denoising of Real Photographs](https://mlsanity.com/paper/1807.04686/) (2018)
- [Zero-Shot Semantic Segmentation](https://mlsanity.com/paper/1906.00817/) (2019)
- [Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Art](https://mlsanity.com/paper/1704.05519/) (2017)
- [SNE-RoadSeg: Incorporating Surface Normal Information into Semantic Segmentation for Accurate Freespace Detection](https://mlsanity.com/paper/2008.11351/) (2020)
- [Adversarial Spatio-Temporal Learning for Video Deblurring](https://mlsanity.com/paper/1804.00533/) (2018)
- [SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving](https://mlsanity.com/paper/1612.01051/) (2016)
- [NO Need to Worry about Adversarial Examples in Object Detection in Autonomous Vehicles](https://mlsanity.com/paper/1707.03501/) (2017)
- [Learning to Generate Images with Perceptual Similarity Metrics](https://mlsanity.com/paper/1511.06409/) (2015)
- [Securing Connected & Autonomous Vehicles: Challenges Posed by Adversarial Machine Learning and The Way Forward](https://mlsanity.com/paper/1905.12762/) (2019)
- [Spatio-temporal video autoencoder with differentiable memory](https://mlsanity.com/paper/1511.06309/) (2015)
- [Image Restoration Using Convolutional Auto-encoders with Symmetric Skip Connections](https://mlsanity.com/paper/1606.08921/) (2016)
- [Adversarial Scene Editing: Automatic Object Removal from Weak Supervision](https://mlsanity.com/paper/1806.01911/) (2018)
- [We Learn Better Road Pothole Detection: from Attention Aggregation to Adversarial Domain Adaptation](https://mlsanity.com/paper/2008.06840/) (2020)
- [MobilBye: Attacking ADAS with Camera Spoofing](https://mlsanity.com/paper/1906.09765/) (2019)
- [Talk Proposal: Towards the Realistic Evaluation of Evasion Attacks using CARLA](https://mlsanity.com/paper/1904.12622/) (2019)

## Built on by (most cited first, 1 of 1)

- [AutoJoin: Efficient Adversarial Training against Gradient-Free Perturbations for Robust Maneuvering via Denoising Autoencoder and Joint Learning](https://mlsanity.com/paper/2205.10933/) (2022)

## Abstract (arXiv)

Autonomous vehicles increasingly rely on cameras to provide the input for perception and scene understanding and the ability of these models to classify their environment and objects, under adverse conditions and image noise is crucial. When the input is, either unintentionally or through targeted attacks, deteriorated, the reliability of autonomous vehicle is compromised. In order to mitigate such phenomena, we propose DriveGuard, a lightweight spatio-temporal autoencoder, as a solution to robustify the image segmentation process for autonomous vehicles. By first processing camera images with DriveGuard, we offer a more universal solution than having to re-train each perception model with noisy input. We explore the space of different autoencoder architectures and evaluate them on a diverse dataset created with real and synthetic images demonstrating that by exploiting spatio-temporal information combined with multi-component loss we significantly increase robustness against adverse image effects reaching within 5-6% of that of the original model on clean images.