Simulating Photo-realistic Snow and Fog on Existing Images for Enhanced CNN Training and Evaluation

Bernuth, Alexander von; Volk, Georg; Bringmann, Oliver

doi:10.1109/itsc.2019.8917367

Cited by 36 publications

(25 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regarding snowy weather, Bernuth et al [167] apply an adverse weather augmentation approach to reuse existing, well-organised, and labelled datasets, KITTI [37] and Cityscapes [39]. The approach produces lifelike and physically correct images to be added to existing ready-touse datasets created with real-world images.…”

Section: ) Self-driving Scenariosmentioning

confidence: 99%

Toward Performing Image Classification and Object Detection With Convolutional Neural Networks in Autonomous Driving Systems: A Survey

Turay

Vladimirova

2022

IEEE Access

View full text Add to dashboard Cite

show abstract

Section: ) Self-driving Scenariosmentioning

confidence: 99%

Toward Performing Image Classification and Object Detection With Convolutional Neural Networks in Autonomous Driving Systems: A Survey

Turay

Vladimirova

2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Thus, a natural course of action is to employ realistic data augmentation pipelines. A lot of research has been done to augment camera images with rain, fog and snow as shown in [1], [2], [3], [4]. However, to our knowledge, realistic data augmentation pipelines to simulate different weather conditions on the LiDAR data are less known.…”

Section: Augmenting Weather Conditionsmentioning

confidence: 99%

“…Thus, there has been a lot of research regarding data augmentation e.g. [1], [2], [3], [4] and evaluating object detection architectures on these augmented datasets. To the best of our knowledge, no current research provides a large scale study on robustness of present 2D and 3D object detectors in degrading weather conditions with data captured in real scenarios.…”

Section: Introductionmentioning

confidence: 99%

Robustness of Object Detectors in Degrading Weather Conditions

Mirza¹,

Buerkle²,

Jarquin³

et al. 2021

Preprint

View full text Add to dashboard Cite

State-of-the-art object detection systems for autonomous driving achieve promising results in clear weather conditions. However, such autonomous safety critical systems also need to work in degrading weather conditions, such as rain, fog and snow. Unfortunately, most approaches evaluate only on the KITTI dataset, which consists only of clear weather scenes. In this paper we address this issue and perform one of the most detailed evaluation on single and dual modality architectures on data captured in real weather conditions. We analyse the performance degradation of these architectures in degrading weather conditions. We demonstrate that an object detection architecture performing good in clear weather might not be able to handle degrading weather conditions. We also perform ablation studies on the dual modality architectures and show their limitations.• The performance should degrade gracefully with changing/degrading weather conditions.

show abstract

“…There are also works on the robustness of object detectors. Image corruptions such as pixel noise, blur, varying weather conditions lead to significant performance drops in object detection models [19,4]. Object detectors are not robust against scale variations [22].…”

Section: Related Workmentioning

confidence: 99%

Shift Equivariance in Object Detection

Manfredi

Wang

2020

Preprint

View full text Add to dashboard Cite

Robustness to small image translations is a highly desirable property for object detectors. However, recent works have shown that CNN-based classifiers are not shift invariant. It is unclear to what extent this could impact object detection, mainly because of the architectural differences between the two and the dimensionality of the prediction space of modern detectors. To assess shift equivariance of object detection models end-to-end, in this paper we propose an evaluation metric, built upon a greedy search of the lower and upper bounds of the mean average precision on a shifted image set. Our new metric shows that modern object detection architectures, no matter if one-stage or two-stage, anchor-based or anchor-free, are sensitive to even one pixel shift to the input images. Furthermore, we investigate several possible solutions to this problem, both taken from the literature and newly proposed, quantifying the effectiveness of each one with the suggested metric. Our results indicate that none of these methods can provide full shift equivariance. Measuring and analyzing the extent of shift variance of different models and the contributions of possible factors, is a first step towards being able to devise methods that mitigate or even leverage such variabilities.

show abstract

Simulating Photo-realistic Snow and Fog on Existing Images for Enhanced CNN Training and Evaluation

Cited by 36 publications

References 22 publications

Toward Performing Image Classification and Object Detection With Convolutional Neural Networks in Autonomous Driving Systems: A Survey

Toward Performing Image Classification and Object Detection With Convolutional Neural Networks in Autonomous Driving Systems: A Survey

Robustness of Object Detectors in Degrading Weather Conditions

Shift Equivariance in Object Detection

Contact Info

Product

Resources

About