Measuring and Modeling Uncertainty Degree for Monocular Depth Estimation
Not on the map. Why?
Abstract (arXiv)
Effectively measuring and modeling the reliability of a trained model is essential to the real-world deployment of monocular depth estimation (MDE) models. However, the intrinsic ill-posedness and ordinal-sensitive nature of MDE pose major challenges to the estimation of uncertainty degree of the trained models. On the one hand, utilizing current uncertainty modeling methods may increase memory consumption and are usually time-consuming. On the other hand, measuring the uncertainty based on model accuracy can also be problematic, where uncertainty reliability and prediction accuracy are not well decoupled. In this paper, we propose to model the uncertainty of MDE models from the perspective of the inherent probability distributions originating from the depth probability volume and its extensions, and to assess it more fairly with more comprehensive metrics. By simply introducing additional training regularization terms, our model, with surprisingly simple formations and without requiring extra modules or multiple inferences, can provide uncertainty estimations with state-of-the-art reliability, and can be further improved when combined with ensemble or sampling methods. A series of experiments demonstrate the effectiveness of our methods.
Key facts. Cited by 1 paper in the map (data through 13 Oct 2026). Builds on 69 of its 103 references that are in the map. First posted on arXiv 19 Jul 2023.
Builds on
69 of 103 references.
References · 69
- 2015Deep Residual Learning for Image Recognitioncited by 51,950
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,043
- 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,111
- 2016Densely Connected Convolutional Networkscited by 6,707
- 2017On Calibration of Modern Neural Networkscited by 4,902
- 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,430
- 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,902
- 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,371
- 2017What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?cited by 2,340
- 2021Vision Transformers for Dense Predictioncited by 1,795
- 2021Plenoxels: Radiance Fields without Neural Networkscited by 1,610
- 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,490
38 more
- 2017Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semanticscited by 1,348
- 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,278
- 2014Depth Map Prediction from a Single Image using a Multi-Scale Deep Networkcited by 1,269
- 2017Unsupervised Learning of Depth and Ego-Motion from Videocited by 779
- 2018Deep Ordinal Regression Network for Monocular Depth Estimationcited by 722
- 2020AdaBins: Depth Estimation using Adaptive Binscited by 598
- 2015Weight Uncertainty in Neural Networkscited by 554
- 2018Pyramid Stereo Matching Networkcited by 489
- 2014Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-Scale Convolutional Architecturecited by 483
- 2018Accurate Uncertainties for Deep Learning Using Calibrated Regressioncited by 422
- 2021Revisiting the Calibration of Modern Neural Networkscited by 388
- 2019From Big to Small: Multi-Scale Local Planar Guidance for Monocular Depth Estimationcited by 380
- 2016Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescuecited by 368
- 2017End-to-End Learning of Geometry and Context for Deep Stereo Regressioncited by 360
- 2021On the Importance of Gradients for Detecting Distributional Shifts in the Wildcited by 339
- 2018DeepMVS: Learning Multi-view Stereopsiscited by 276
- 2015Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understandingcited by 218
- 2020Learning to Recover 3D Scene Shape from a Single Imagecited by 214
- 2019Addressing Failure Prediction by Learning Model Confidencecited by 185
- 2020On the uncertainty of self-supervised monocular depth estimationcited by 182
- 2016Single-Image Depth Perception in the Wildcited by 179
- 2022BinsFormer: Revisiting Adaptive Bins for Monocular Depth Estimationcited by 167
- 2015Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inferencecited by 159
- 2021CFNet: Cascade and Fused Cost Volume for Robust Stereo Matchingcited by 154
- 2022NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth Estimationcited by 151
- 2018Uncertainty Estimates and Multi-Hypotheses Networks for Optical Flowcited by 135
- 2018High Quality Monocular Depth Estimation via Transfer Learningcited by 133
- 2021TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformerscited by 117
- 2022DepthFormer: Exploiting Long-Range Correlation and Local Information for Accurate Monocular Depth Estimationcited by 116
- 2020Confidence-Aware Learning for Deep Neural Networkscited by 113
- 2021DEUP: Direct Epistemic Uncertainty Predictioncited by 102
- 2019Distribution Calibration for Regressioncited by 89
- 2019Evaluating and Calibrating Uncertainty Prediction in Regression Taskscited by 85
- 2022Rethinking Depth Estimation for Multi-View Stereo: A Unified Representationcited by 82
- 2020Dense Hybrid Recurrent Multi-view Stereo Net with Dynamic Consistency Checkingcited by 81
- 2019Neural RGB->D Sensing: Depth and Uncertainty from a Video Cameracited by 69
- 2019Adaptive Unimodal Cost Volume Filtering for Deep Stereo Matchingcited by 55
- 2020Individual Calibration with Randomized Forecastingcited by 50
50 most cited of 69 All 69 (JSON)
Built on by
Citing papers per year: 2026: 1.
1 paper.