VOS: Learning What You Don't Know by Virtual Outlier Synthesis

Du, Xu; Wang, Zhaoning; Cai, Mu; Li, Yixuan

doi:10.48550/arxiv.2202.01197

Cited by 16 publications

(31 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We show the performance of RODD in Tables 1 and 2 for CIFAR-10 and CIFAR-100, respectively. Our method achieves an FPR95 improvement of 21.66%, compared to the most recently reported SOTA [6], on CIFAR-10. We obtain similar performance gains for CIFAR-100 dataset as well.…”

Section: Resultsmentioning

confidence: 69%

“…As in [6,29], the OOD detection performance of RODD is evaluated using the following metrics: (i) FPR95 indicates the false positive rate (FPR) at 95% true positive rate (TPR) and (ii) AUROC, which is defined as the Area Under the Receiver Operating Characteristic curve. As RODD is a probabilistic approach, sampling is preformed on the ID and OOD data during the test time to ensure the probabilistic settings.…”

Section: Evaluation Metrics and Inference Criterionmentioning

confidence: 99%

“…In the realworld scenarios, OOD detectors are distribution-agnostic. To overcome this limitation, some other methods that are independent of OOD data during the training process have been proposed [6,13,14,29,34,37]. Such methods either use the membership probabilities [6,13,14,29] or a feature embedding [34,37] to calculate an uncertainty score.…”

Section: Introductionmentioning

confidence: 99%

“…To overcome this limitation, some other methods that are independent of OOD data during the training process have been proposed [6,13,14,29,34,37]. Such methods either use the membership probabilities [6,13,14,29] or a feature embedding [34,37] to calculate an uncertainty score. In [34], the authors proposed to reconstruct the samples to produce a discriminate feature space.…”

Section: Introductionmentioning

confidence: 99%

“…In [34], the authors proposed to reconstruct the samples to produce a discriminate feature space. Similarly, [6] proposed synthesizing virtual outliers to regularize the model's decision boundary during training. Nevertheless, the performance of the methods that rely on either reconstruction or generation [6,26,34] degrades on large-scale datasets or video classification scenarios.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection

Khalid¹,

Esmaeili²,

Karim³

et al. 2022

Preprint

View full text Add to dashboard Cite

Recent studies have started to address the concern of detecting and rejecting the out-of-distribution (OOD) samples as a major challenge in the safe deployment of deep learning (DL) models. It is desired that the DL model should only be confident about the in-distribution (ID) data which reinforces the driving principle of the OOD detection. In this paper, we propose a simple yet effective generalized OOD detection method independent of out-of-distribution datasets. Our approach relies on self-supervised feature learning of the training samples, where the embeddings lie on a compact low-dimensional space. Motivated by the recent studies that show self-supervised adversarial contrastive learning helps robustify the model, we empirically show that a pre-trained model with self-supervised contrastive learning yields a better model for uni-dimensional feature learning in the latent space. The method proposed in this work, referred to as RODD, outperforms SOTA detection performance on extensive suite of benchmark datasets on OOD detection tasks. On the CIFAR-100 benchmarks, RODD achieves a 26.97 % lower false positive rate (FPR@95) compared to SOTA methods. Our code is publicly available. 1

show abstract

Section: Resultsmentioning

confidence: 69%

Section: Evaluation Metrics and Inference Criterionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection

Khalid¹,

Esmaeili²,

Karim³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

Bolstering Maritime Object Detection with Synthetic Data

et al. 2022

View full text Add to dashboard Cite

Robust 3D Object Detection in Cold Weather Conditions

Piroli

Dallabetta

Walessa

et al. 2022

2022 IEEE Intelligent Vehicles Symposium (IV)

View full text Add to dashboard Cite

LiDAR sensors are used in autonomous driving applications to accurately perceive the environment. However, they are affected by adverse weather conditions such as snow, fog, and rain. These everyday phenomena introduce unwanted noise into the measurements, severely degrading the performance of LiDAR-based perception systems. In this work, we propose a framework for improving the robustness of LiDARbased 3D object detectors against road spray. Our approach uses a state-of-the-art adverse weather detection network to filter out spray from the LiDAR point cloud, which is then used as input for the object detector. In this way, the detected objects are less affected by the adverse weather in the scene, resulting in a more accurate perception of the environment. In addition to adverse weather filtering, we explore the use of radar targets to further filter false positive detections. Tests on real-world data show that our approach improves the robustness to road spray of several popular 3D object detectors.

show abstract

VOS: Learning What You Don't Know by Virtual Outlier Synthesis

Cited by 16 publications

References 24 publications

RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection

RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection

Bolstering Maritime Object Detection with Synthetic Data

Robust 3D Object Detection in Cold Weather Conditions

Contact Info

Product

Resources

About