Recent self-supervised clustering-based pre-training techniques like DINO and CrIBo have shown impressive results for downstream detection and segmentation tasks. However, real-world applications such as autonomous driving face challenges with imbalanced object class and size distributions and complex scene geometries. In this paper, we propose S3PT a novel scene semantics and structure guided clustering to provide more scene-consistent objectives for self-supervised training. Specifically, our contributions are threefold: First, we incorporate semantic distribution consistent clustering to encourage better representation of rare classes such as motorcycles or animals. Second, we introduce object diversity consistent spatial clustering, to handle imbalanced and diverse object sizes, ranging from large background areas to small objects such as pedestrians and traffic signs. Third, we propose a depth-guided spatial clustering to regularize learning based on geometric information of the scene, thus further refining region separation on the feature level. Our learned representations significantly improve performance in downstream semantic segmentation and 3D object detection tasks on the nuScenes, nuImages, and Cityscapes datasets and show promising domain translation properties.

LiDAR is crucial for robust 3D scene perception in autonomous driving. LiDAR perception has the largest body of literature after camera perception. However, multi-task learning across tasks like detection, segmentation, and motion estimation using LiDAR remains relatively unexplored, especially on automotive-grade embedded platforms. We present a real-time multi-task convolutional neural network for LiDAR-based object detection, semantics, and motion segmentation. The unified architecture comprises a shared encoder and task-specific decoders, enabling joint representation learning. We propose a novel Semantic Weighting and Guidance (SWAG) module to transfer semantic features for improved object detection selectively. Our heterogeneous training scheme combines diverse datasets and exploits complementary cues between tasks. The work provides the first embedded implementation unifying these key perception tasks from LiDAR point clouds achieving 3ms latency on the embedded NVIDIA Xavier platform. We achieve state-of-the-art results for two tasks, semantic and motion segmentation, and close to state-of-the-art performance for 3D object detection. By maximizing hardware efficiency and leveraging multi-task synergies, our method delivers an accurate and efficient solution tailored for real-world automated driving deployment. Qualitative results can be seen at https://youtu.be/H-hWRzv2lIY.

Semantic scene segmentation from a bird’s-eye-view (BEV) perspective plays a crucial role in facilitating planning and decision-making for mobile robots. Although recent vision-only methods have demonstrated notable advancements in performance, they often struggle under adverse illumination conditions such as rain or nighttime. While active sensors offer a solution to this challenge, the prohibitively high cost of LiDARs remains a limiting factor. Fusing camera data with automotive radars poses a more inexpensive alternative but has received less attention in prior research. In this work, we aim to advance this promising avenue by introducing BEVCar, a novel approach for joint BEV object and map segmentation. The core novelty of our approach lies in first learning a point-based encoding of raw radar data, which is then leveraged to efficiently initialize the lifting of image features into the BEV space. We perform extensive experiments on the nuScenes dataset and demonstrate that BEVCar outperforms the current state of the art. Moreover, we show that incorporating radar information significantly enhances robustness in challenging environmental conditions and improves segmentation performance for distant objects. To foster future research, we provide the weather split of the nuScenes dataset used in our experiments, along with our code and trained models at http://bevcar.cs.uni-freiburg.de.

Recent advances in 3D object detection (3DOD) have obtained remarkably strong results for LiDAR-based models. In contrast, surround-view 3DOD models based on multiple camera images underperform due to the necessary view transformation of features from perspective view (PV) to a 3D world representation which is ambiguous due to missing depth information. This paper introduces X3KD, a comprehensive knowledge distillation framework across different modalities, tasks, and stages for multi-camera 3DOD. Specifically, we propose cross-task distillation from an instance segmentation teacher (X-IS) in the PV feature extraction stage providing supervision without ambiguous error backpropagation through the view transformation. After the transformation, we apply cross-modal feature distillation (X-FD) and adversarial training (X-AT) to improve the 3D world representation of multi-camera features through the information contained in a LiDAR-based 3DOD teacher. Finally, we also employ this teacher for cross-modal output distillation (X-OD), providing dense supervision at the prediction stage. We perform extensive ablations of knowledge distillation at different stages of multi-camera 3DOD. Our final X3KD model outperforms previous state-of-the-art approaches on the nuScenes and Waymo datasets and generalizes to RADAR-based 3DOD. Qualitative results video at https://youtu.be/1do9DPFmr38.

Bird’s-eye-view (BEV) grid is a typical representation of the perception of road components, e.g., drivable area, in autonomous driving. Most existing approaches rely on cameras only to perform segmentation in BEV space, which is fundamentally constrained by the absence of reliable depth information. The latest works leverage both camera and LiDAR modalities but suboptimally fuse their features using simple, concatenation-based mechanisms.In this paper, we address these problems by enhancing the alignment of the unimodal features in order to aid feature fusion, as well as enhancing the alignment between the cameras’ perspective view (PV) and BEV representations. We propose X-Align, a novel end-to-end cross-modal and cross-view learning framework for BEV segmentation consisting of the following components: (i) a novel CrossModal Feature Alignment (X-FA) loss, (ii) an attentionbased Cross-Modal Feature Fusion (X-FF) module to align multi-modal BEV features implicitly, and (iii) an auxiliary PV segmentation branch with Cross-View Segmentation Alignment (X-SA) losses to improve the PV-to-BEV transformation. We evaluate our proposed method across two commonly used benchmark datasets, i.e., nuScenes and KITTI-360. Notably, X-Align significantly outperforms the state-of-the-art by 3 absolute mIoU points on nuScenes. We also provide extensive ablation studies to demonstrate the effectiveness of the individual components.

Surround-view fisheye cameras are commonly used for near-field sensing in automated driving. Four fisheye cameras on four sides of the vehicle are sufficient to cover 360° around the vehicle capturing the entire near-field region. Some primary use cases are automated parking, traffic jam assist, and urban driving. There are limited datasets and very little work on near-field perception tasks as the focus in automotive perception is on far-field perception. In contrast to far-field, surround-view perception poses additional challenges due to high precision object detection requirements of 10cm and partial visibility of objects. Due to the large radial distortion of fisheye cameras, standard algorithms cannot be extended easily to the surround-view use case. Thus, we are motivated to provide a self-contained reference for automotive fisheye camera perception for researchers and practitioners. Firstly, we provide a unified and taxonomic treatment of commonly used fisheye camera models. Secondly, we discuss various perception tasks and existing literature. Finally, we discuss the challenges and future direction.

Generating a detailed near-field perceptual model of the environment is an important and challenging problem in both self-driving vehicles and autonomous mobile robotics. A Bird's Eye View (BEV) map, providing a panoptic representation, is a commonly used approach that provides a simplified 2D representation of the vehicle's surroundings with accurate semantic level segmentation for many downstream tasks. Current state-of-the art approaches to generate BEV-maps employ a Convolutional Neural Network (CNN) backbone to create feature-maps which are passed through a spatial transformer to project the derived features onto the BEV coordinate frame. In this paper, we evaluate the use of vision transformers (ViT) as a backbone architecture to generate BEV maps. Our network architecture, ViT-BEVSeg, employs standard vision transformers to generate a multi-scale representation of the input image. The resulting representation is then provided as an input to a spatial transformer decoder module which outputs segmentation maps in the BEV grid. We evaluate our approach on the nuScenes dataset demonstrating a considerable improvement in the performance relative to state-of-the-art approaches. Code is available at https://github.com/robotvisionmu/ViT-BEVSeg.

While deep neural networks (DNNs) achieve impressive performance on environment perception tasks, their sensitivity to adversarial perturbations limits their use in practical applications. In this paper, we (i) propose a novel adversarial perturbation detection scheme based on multi-task perception of complex vision tasks (i.e., depth estimation and semantic segmentation). Specifically, adversarial perturbations are detected by inconsistencies between extracted edges of the input image, the depth output, and the segmentation output. To further improve this technique, we (ii) develop a novel edge consistency loss between all three modalities, thereby improving their initial consistency which in turn supports our detection scheme. We verify our detection scheme's effectiveness by employing various known attacks and image noises. In addition, we (iii) develop a multi-task adversarial attack, aiming at fooling both tasks as well as our detection scheme. Experimental evaluation on the Cityscapes and KITTI datasets shows that under an assumption of a 5% false positive rate up to 100% of images are correctly detected as adversarially perturbed, depending on the strength of the perturbation. Code is available at https://github.com/ifnspaml/AdvAttackDet. A short video at https://youtu.be/KKa6gOyWmH4 provides qualitative results.

Electric Vehicles are increasingly common, with inductive chargepads being considered a convenient and efficient means of charging electric vehicles. However, drivers are typically poor at aligning the vehicle to the necessary accuracy for efficient inductive charging, making the automated alignment of the two charging plates desirable. In parallel to the electrification of the vehicular fleet, automated parking systems that make use of surround-view camera systems are becoming increasingly popular. In this work, we propose a system based on the surround-view camera architecture to detect, localize, and automatically align the vehicle with the inductive chargepad. The visual design of the chargepads is not standardized and not necessarily known beforehand. Therefore, a system that relies on offline training will fail in some situations. Thus, we propose a self-supervised online learning method that leverages the driver’s actions when manually aligning the vehicle with the chargepad and combine it with weak supervision from semantic segmentation and depth to learn a classifier to auto-annotate the chargepad in the video for further training. In this way, when faced with a previously unseen chargepad, the driver needs only manually align the vehicle a single time. As the chargepad is flat on the ground, it is not easy to detect it from a distance. Thus, we propose using a Visual SLAM pipeline to learn landmarks relative to the chargepad to enable alignment from a greater range. We demonstrate the working system on an automated vehicle as illustrated in the videohttps://youtu.be/_cLCmkW4UYo. To encourage further research, we will share a chargepad dataset used in this work (an initial version of the dataset is sharedhttps://drive.google.com/drive/folders/1KeLFIqOnhU2CGsD0vbiN9UqKmBSyHERdhere).

Pedestrian Detection is the most critical module of an Autonomous Driving system. Although a camera is commonly used for this purpose, its quality degrades severely in low-light night time driving scenarios. On the other hand, the quality of a thermal camera image remains unaffected in similar conditions. This paper proposes an end-to-end multimodal fusion model for pedestrian detection using RGB and thermal images. Its novel spatio-contextual deep network architecture is capable of exploiting the multimodal input efficiently. It consists of two distinct deformable ResNeXt-50 encoders for feature extraction from the two modalities. Fusion of these two encoded features takes place inside a multimodal feature embedding module (MuFEm) consisting of several groups of a pair of Graph Attention Network and a feature fusion unit. The output of the last feature fusion unit of MuFEm is subsequently passed to two CRFs for their spatial refinement. Further enhancement of the features is achieved by applying channel-wise attention and extraction of contextual information with the help of four RNNs traversing in four different directions. Finally, these feature maps are used by a single-stage decoder to generate the bounding box of each pedestrian and the score map. We have performed extensive experiments of the proposed framework on three publicly available multimodal pedestrian detection benchmark datasets, namely KAIST, CVC-14, and UTokyo. The results on each of them improved the respective state-of-the-art performance. A short video giving an overview of this work along with its qualitative results can be seen athttps://youtu.be/FDJdSifuuCs. Our source code will be released upon publication of the paper.

Cameras are the primary sensor in automated driving systems. They provide high information density and are optimal for detecting road infrastructure cues laid out for human vision. Surround-view camera systems typically comprise of four fisheye cameras with 190°+ field of view covering the entire 360° around the vehicle focused on near-field sensing. They are the principal sensors for low-speed, high accuracy, and close-range sensing applications, such as automated parking, traffic jam assistance, and low-speed emergency braking. In this work, we provide a detailed survey of such vision systems, setting up the survey in the context of an architecture that can be decomposed into four modular components namely Recognition, Reconstruction, Relocalization, and Reorganization. We jointly call this the4R Architecture. We discuss how each component accomplishes a specific aspect and provide a positional argument that they can be synergized to form a complete perception system for low-speed automation. We support this argument by presenting results from previous works and by presenting architecture proposals for such a system. Qualitative results are presented in the video athttps://youtu.be/ae8bCOF77uY.

With the development of deep representation learning, the domain of reinforcement learning (RL) has become a powerful learning framework now capable of learning complex policies in high dimensional environments. This review summarises deep reinforcement learning (DRL) algorithms and provides a taxonomy of automated driving tasks where (D)RL methods have been employed, while addressing key computational challenges in real world deployment of autonomous driving agents. It also delineates adjacent domains such as behavior cloning, imitation learning, inverse reinforcement learning that are related but are not classical RL algorithms. The role of simulators in training agents, methods to validate, test and robustify existing solutions in RL are discussed.

A 360° perception of scene geometry is essential for automated driving, notably for parking and urban driving scenarios. Typically, it is achieved using surround-view fisheye cameras, focusing on the near-field area around the vehicle. The majority of current depth estimation approaches focus on employing just a single camera, which cannot be straightforwardly generalized to multiple cameras. The depth estimation model must be tested on a variety of cameras equipped to millions of cars with varying camera geometries. Even within a single car, intrinsics vary due to manufacturing tolerances. Deep learning models are sensitive to these changes, and it is practically infeasible to train and test on each camera variant. As a result, we present novel camera-geometry adaptive multi-scale convolutions which utilize the camera parameters as a conditional input, enabling the model to generalize to previously unseen fisheye cameras. Additionally, we improve the distance estimation by pairwise and patchwise vector-based self-attention encoder networks. We evaluate our approach on the Fisheye WoodScape surround-view dataset, significantly improving over previous approaches. We also show a generalization of our approach across different camera viewing angles and perform extensive experiments to support our contributions. To enable comparison with other approaches, we evaluate the front camera data on the KITTI dataset (pinhole camera images) and achieve state-of-the-art performance among self-supervised monocular methods. An overview video with qualitative results is provided athttps://youtu.be/bmX0UcU9wtA. Baseline code and dataset will be made public.1

Panoptic Segmentation aims to provide an understanding of background (stuff) and instances of objects (things) at a pixel level. It combines the separate tasks of semantic segmentation (pixel level classification) and instance segmentation to build a single unified scene understanding task. Typically, panoptic segmentation is derived by combining semantic and instance segmentation tasks that are learned separately or jointly (multi-task networks). In general, instance segmentation networks are built by adding a foreground mask estimation layer on top of object detectors or using instance clustering methods that assign a pixel to an instance center. In this work, we present a fully convolution neural network that learns instance segmentation from semantic segmentation and instance contours (boundaries of things). Instance contours along with semantic segmentation yield a boundary aware semantic segmentation of things. Connected component labeling on these results produces instance segmentation. We merge semantic and instance segmentation results to output panoptic segmentation. We evaluate our proposed method on the CityScapes dataset to demonstrate qualitative and quantitative performances along with several ablation studies. Our overview video can be accessed from https://youtu.be/wBtcxRhG3e0.

Moving object segmentation is a crucial task for autonomous vehicles as it can be used to segment objects in a class agnostic manner based on their motion cues. It enables the detection of unseen objects during training (e.g., moose or a construction truck) based on their motion and independent of their appearance. Although pixel-wise motion segmentation has been studied in autonomous driving literature, it has been rarely addressed at the instance level, which would help separate connected segments of moving objects leading to better trajectory planning. As the main issue is the lack of large public datasets, we create a new InstanceMotSeg dataset comprising of 12.9K samples improving upon our KITTIMoSeg dataset. In addition to providing instance level annotations, we have added 4 additional classes which is crucial for studying class agnostic motion segmentation. We adapt YOLACT and implement a motion-based class agnostic instance segmentation model which would act as a baseline for the dataset. We also extend it to an efficient multi-task model which additionally provides semantic instance segmentation sharing the encoder. The model then learns separate prototype coefficients within the class agnostic and semantic heads providing two independent paths of object detection for redundant safety. To obtain real-time performance, we study different efficient encoders and obtain 39 fps on a Titan Xp GPU using MobileNetV2 with an improvement of 10% mAP relative to the baseline. Our model improves the previous state of the art motion segmentation method by 3.3%. The dataset and qualitative results video are shared in our website at https://sites.google.com/view/instancemotseg.

State-of-the-art self-supervised learning approaches for monocular depth estimation usually suffer from scale ambiguity. They do not generalize well when applied on distance estimation for complex projection models such as in fisheye and omnidirectional cameras. This paper introduces a novel multi-task learning strategy to improve self-supervised monocular distance estimation on fisheye and pinhole camera images. Our contribution to this work is threefold: Firstly, we introduce a novel distance estimation network architecture using a self-attention based encoder coupled with robust semantic feature guidance to the decoder that can be trained in a one-stage fashion. Secondly, we integrate a generalized robust loss function, which improves performance significantly while removing the need for hyperparameter tuning with the reprojection loss. Finally, we reduce the artifacts caused by dynamic objects violating static world assumptions using a semantic masking strategy. We significantly improve upon the RMSE of previous work on fisheye by 25% reduction in RMSE. As there is little work on fisheye cameras, we evaluated the proposed method on KITTI using a pinhole model. We achieved state-of-the-art performance among self-supervised methods without requiring an external scale estimation.

Object detection is a comprehensively studied problem in autonomous driving. However, it has been relatively less explored in the case of fisheye cameras. The standard bounding box fails in fisheye cameras due to the strong radial distortion, particularly in the image's periphery. We explore better representations like oriented bounding box, ellipse, and generic polygon for object detection in fisheye images in this work. We use the IoU metric to compare these representations using accurate instance segmentation ground truth. We design a novel curved bounding box model that has optimal properties for fisheye distortion models. We also design a curvature adaptive perimeter sampling method for obtaining polygon vertices, improving relative mAP score by 4.9% compared to uniform sampling. Overall, the proposed polygon model improves mIoU relative accuracy by 40.3%. It is the first detailed study on object detection on fisheye cameras for autonomous driving scenarios to the best of our knowledge. The dataset1comprising of 10,000 images along with all the object representations ground truth will be made public to encourage further research. We summarize our work in a short video with qualitative results at https://youtu.be/iLkOzvJpL-A.

Wide-angle fisheye cameras are commonly used in automated driving for parking and low-speed navigation tasks. Four of such cameras form a surround-view system that provides a complete and detailed view of the vehicle. These cameras are directly exposed to harsh environmental settings and can get soiled very easily by mud, dust, water, frost. Soiling on the camera lens can severely degrade the visual perception algorithms, and a camera cleaning system triggered by a soiling detection algorithm is increasingly being deployed. While adverse weather conditions, such as rain, are getting attention recently, there is only limited work on general soiling. The main reason is the difficulty in collecting a diverse dataset as it is a relatively rare event.We propose a novel GAN based algorithm for generating unseen patterns of soiled images. Additionally, the proposed method automatically provides the corresponding soiling masks eliminating the manual annotation cost. Augmentation of the generated soiled images for training improves the accuracy of soiling detection tasks significantly by 18% demonstrating its usefulness. The manually annotated soiling dataset and the generated augmentation dataset will be made public. We demonstrate the generalization of our fisheye trained GAN model on the Cityscapes dataset. We provide an empirical evaluation of the degradation of the semantic segmentation algorithm with the soiled data.

Fisheye cameras are commonly used in applications like autonomous driving and surveillance to provide a large field of view (> 180o). However, they come at the cost of strong non-linear distortions which require more complex algorithms. In this paper, we explore Euclidean distance estimation on fisheye cameras for automotive scenes. Obtaining accurate and dense depth supervision is difficult in practice, but self-supervised learning approaches show promising results and could potentially overcome the problem. We present a novel self-supervised scale-aware framework for learning Euclidean distance and ego-motion from raw monocular fisheye videos without applying rectification. While it is possible to perform piece-wise linear approximation of fisheye projection surface and apply standard rectilinear models, it has its own set of issues like re-sampling distortion and discontinuities in transition regions. To encourage further research in this area, we will release our dataset as part of the WoodScape project [1]. We further evaluated the proposed algorithm on the KITTI dataset and obtained state-of-the-art results comparable to other self-supervised monocular methods. Qualitative results on an unseen fisheye video demonstrate impressive performance1.

In classical computer vision, rectification is an integral part of multi-view depth estimation. It typically includes epipolar rectification and lens distortion correction. This process simplifies the depth estimation significantly, and thus it has been adopted in CNN approaches. However, rectification has several side effects, including a reduced field of view (FOV), resampling distortion, and sensitivity to calibration errors. The effects are particularly pronounced in case of significant distortion (e.g., wide-angle fisheye cameras). In this paper, we propose a generic scale-aware self-supervised pipeline for estimating depth, euclidean distance, and visual odometry from unrectified monocular videos. We demonstrate a similar level of precision on the unrectified KITTI dataset with barrel distortion comparable to the rectified KITTI dataset. The intuition being that the rectification step can be implicitly absorbed within the CNN model, which learns the distortion model without increasing complexity. Our approach does not suffer from a reduced field of view and avoids computational costs for rectification at inference time. To further illustrate the general applicability of the proposed framework, we apply it to wide-angle fisheye cameras with 190° horizontal field of view. The training framework UnRectDepthNet takes in the camera distortion model as an argument and adapts projection and unprojection functions accordingly. The proposed algorithm is evaluated further on the KITTI rectified dataset, and we achieve state-of-the-art results that improve upon our previous work FisheyeDistanceNet [1]. Qualitative results on a distorted test scene video sequence indicate excellent performance1.

Fisheye cameras are commonly employed for obtaining a large field of view in surveillance, augmented reality and in particular automotive applications. In spite of their prevalence, there are few public datasets for detailed evaluation of computer vision algorithms on fisheye images. We release the first extensive fisheye automotive dataset, WoodScape, named after Robert Wood who invented the fisheye camera in 1906. WoodScape comprises of four surround view cameras and nine tasks including segmentation, depth estimation, 3D bounding box detection and soiling detection. Semantic annotation of 40 classes at the instance level is provided for over 10,000 images and annotation for other tasks are provided for over 100,000 images. With WoodScape, we would like to encourage the community to adapt computer vision models for fisheye camera instead of using naive rectification.

High performance ultrasonic sensors are commonly used in medical applications. Ultrasonic sensors are also used in autonomous driving for near-field sensing. Its low cost is one of its main advantages and thus it is necessary to have low computational complexity for the classification algorithms. In this paper, we propose an efficient capsule neural network to provide a detailed height analysis of detected objects. We apply a novel re-sorting and re-shaping method to feed the neural network with ultrasonic data. We evaluated the proposed method under different environmental conditions and obtained high validation accuracy of 99% at a run-time of 0.2 ms. We compare it with a conventional CNN and demonstrate better accuracy and run-time.

Generative Adversarial Networks (GANs) have achieved outstanding results in generation of realistic data, particularly for image data. Autonomous driving systems are equipped with a large suite of sensors to obtain robustness and redundancy. Ultrasonic sensors are commonly used because of their low-cost and reliability of near-field distance estimation. However, machine learning algorithms are not commonly used for ultrasonic data, as it requires extensive datasets whose creation is time-consuming, expensive and inflexible to hardware and environmental changes. On the other hand, there exists no method to simulate these signals deterministically. Thus, we present a novel approach for synthetic ultrasonic signal simulation using conditional GANs (cGANs). To the best of our knowledge, we present the first realistic data augmentation for automotive ultrasonics sensors. The performance of cGANs allows us to bring the environment simulation to a high quality close to realistic data. By using our setup and environmental parameters as condition, the proposed approach is flexible to external influences. Due to its low complexity and smaller time effort needed for data generation, the proposed method outperforms other simulation algorithms such as finite element method. We verify the outstanding accuracy and realism of our method by applying a detailed statistical analysis and comparing the generated data to an extensive amount of measured signals.

Semantic segmentation benefits robotics related applications, especially autonomous driving. Most of the research on semantic segmentation only focuses on increasing the accuracy of segmentation models with little attention to computationally efficient solutions. The few work conducted in this direction does not provide principled methods to evaluate the different design choices for segmentation. In this paper, we address this gap by presenting a real-time semantic segmentation benchmarking framework with a decoupled design for feature extraction and decoding methods. The framework is comprised of different network architectures for feature extraction such as VGG16, Resnet18, MobileNet, and ShuffleNet. It is also comprised of multiple meta-architectures for segmentation that define the decoding methodology. These include SkipNet, UNet, and Dilation Frontend. Experimental results are presented on the Cityscapes dataset for urban scenes. The modular design allows novel architectures to emerge, that lead to 143x GFLOPs reduction in comparison to SegNet. This benchmarking framework is publicly available at11https://github.com/MSiam/TFSegmentation.