Robust Monocular Localization in Sparse HD Maps Leveraging Multi-Task Uncertainty Estimation

Kürsat, Petek,; Sirohi, Kshitij; Büscher, D.; Burgard, Wolfram

doi:10.1109/icra46639.2022.9812266

Cited by 14 publications

(6 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This ambiguity stays unresolved due to a missing additional constraint for β i in the regularizer L R i . Curiously, if applied to synthetic and real-world data, the aforementioned approach does yield reasonable results (Amini et al 2020;Liu et al 2021;Cai et al 2021;Soleimany et al 2021;Singh et al 2022;Petek et al 2022;Li and Liu 2021). In particular, areas with a low data density during training that result in a large model uncertainty during inference are identified with large values in the epistemic uncertainty if estimated according to Eq.…”

Section: State Of the Art And Its Issuesmentioning

confidence: 99%

“…A different approach to uncertainty-aware NNs may be useful to more efficiency quantify, and also to disentangle, the several types of uncertainties: Deep Evidential Regression (DER) aims to simultaneously predict both uncertainty types in a single forward pass without sampling or utilization of out-of-distribution data, based on learning evidential distributions for aleatoric and epistemic uncertainties (Amini et al 2020). Yet only with simple empirical demonstrations on univariate regression tasks, this technique has already been applied and recommended in medical and other safety critical applications (Liu et al 2021;Soleimany et al 2021;Cai et al 2021;Chen, Bromuri, and van Eekelen 2021;Singh et al 2022;Petek et al 2022;Li and Liu 2021). With an alternative derivation and experimentation, we identify theoretical shortcomings that do not justify the empirical results let alone the assumed reliability in practice -it can be vital to understand to what degree the uncertainty estimations are trustworthy.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The Unreasonable Effectiveness of Deep Evidential Regression

Meinert¹,

Gawlikowski

Lavin³

2023

AAAI

View full text Add to dashboard Cite

There is a significant need for principled uncertainty reasoning in machine learning systems as they are increasingly deployed in safety-critical domains. A new approach with uncertainty-aware regression-based neural networks (NNs), based on learning evidential distributions for aleatoric and epistemic uncertainties, shows promise over traditional deterministic methods and typical Bayesian NNs, notably with the capabilities to disentangle aleatoric and epistemic uncertainties. Despite some empirical success of Deep Evidential Regression (DER), there are important gaps in the mathematical foundation that raise the question of why the proposed technique seemingly works. We detail the theoretical shortcomings and analyze the performance on synthetic and real-world data sets, showing that Deep Evidential Regression is a heuristic rather than an exact uncertainty quantification. We go on to discuss corrections and redefinitions of how aleatoric and epistemic uncertainties should be extracted from NNs.

show abstract

Section: State Of the Art And Its Issuesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

The Unreasonable Effectiveness of Deep Evidential Regression

Meinert¹,

Gawlikowski

Lavin³

2023

AAAI

View full text Add to dashboard Cite

show abstract

“…However, these methods do not consider perception uncertainty. Petek et al [23] provided a multi-task perception module with uncertainty estimation in their localization algorithm. They train their network to detect the drivable areas and utilize the perception uncertainty to extract lane boundaries and match them with the lane-based HD map to localize.…”

Section: Perception-uncertainty In Localizationmentioning

confidence: 99%

Uncertainty-Aware Panoptic Segmentation

Sirohi

Marvi

Büscher

et al. 2023

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

The availability of a robust map-based localization system is essential for the operation of many autonomously navigating vehicles. Since uncertainty is an inevitable part of perception, it is beneficial for the robustness of the robot to consider it in typical downstream tasks of navigation stacks. In particular localization and mapping methods, which in modern systems often employ convolutional neural networks (CNNs) for perception tasks, require proper uncertainty estimates. In this work, we present uncertainty-aware Panoptic Localization and Mapping (uPLAM), which employs pixel-wise uncertainty estimates for panoptic CNNs as a bridge to fuse modern perception with classical probabilistic localization and mapping approaches. Beyond the perception, we introduce an uncertainty-based map aggregation technique to create accurate panoptic maps, containing surface semantics and landmark instances. Moreover, we provide cell-wise map uncertainties, and present a particle filterbased localization method that employs perception uncertainties. Extensive evaluations show that our proposed incorporation of uncertainties leads to more accurate maps with reliable uncertainty estimates and improved localization accuracy. Additionally, we present the Freiburg Panoptic Driving dataset for evaluating panoptic mapping and localization methods. We make our code and dataset available at: http://uplam.cs.uni-freiburg.de

show abstract

“…The DER places Normal Inverse-Gamma (N IG) priors on the likelihood function and formulates learning as an evidence acquisition process. Due to only minor modifications to neural networks without the sampling and the ability to quantify both epistemic and aleatoric uncertainties in a single forward pass, DER have gained widespread adoption (Liu et al 2021;Chen, Bromuri, and van Eekelen 2021;Singh et al 2022;Petek et al 2022;Li and Liu 2022;Amini et al 2020;Ma et al 2021;Charpentier et al 2022;Oh and Shin 2022;Pandey and Yu 2023a). Despite the attractive ability for uncertainty quantification, the DER's error is noticeably bigger than standard regres-…”

Section: Introductionmentioning

confidence: 99%

The Evidence Contraction Issue in Deep Evidential Regression: Discussion and Solution

Wu,

Shi,

Dong

et al. 2024

AAAI

View full text Add to dashboard Cite

Deep Evidential Regression (DER) places a prior on the original Gaussian likelihood and treats learning as an evidence acquisition process to quantify uncertainty. For the validity of the evidence theory, DER requires specialized activation functions to ensure that the prior parameters remain non-negative. However, such constraints will trigger evidence contraction, causing sub-optimal performance. In this paper, we analyse DER theoretically, revealing the intrinsic limitations for sub-optimal performance: the non-negativity constraints on the Normal Inverse-Gamma (NIG) prior parameter trigger the evidence contraction under the specialized activation function, which hinders the optimization of DER performance. On this basis, we design a Non-saturating Uncertainty Regularization term, which effectively ensures that the performance is further optimized in the right direction. Experiments on real-world datasets show that our proposed approach improves the performance of DER while maintaining the ability to quantify uncertainty.

show abstract

Robust Monocular Localization in Sparse HD Maps Leveraging Multi-Task Uncertainty Estimation

Cited by 14 publications

References 23 publications

The Unreasonable Effectiveness of Deep Evidential Regression

The Unreasonable Effectiveness of Deep Evidential Regression

Uncertainty-Aware Panoptic Segmentation

The Evidence Contraction Issue in Deep Evidential Regression: Discussion and Solution

Contact Info

Product

Resources

About