Abstract:Single-view depth estimation refers to the ability to derive three-dimensional information per pixel from a single two-dimensional image. Single-view depth estimation is an ill-posed problem because there are multiple depth solutions that explain 3D geometry from a single view. While deep neural networks have been shown to be effective at capturing depth from a single view, the majority of current methodologies are deterministic in nature. Accounting for uncertainty in the predictions can avoid disastrous consequences when applied to fields such as autonomous driving or medical robotics. We have addressed this problem by quantifying the uncertainty of supervised single-view depth for Bayesian deep neural networks. There are scenarios, especially in medicine in the case of endoscopic images, where such annotated data is not available. To alleviate the lack of data, we present a method that improves the transition from synthetic to real domain methods. We introduce an uncertainty-aware teacher-student architecture that is trained in a self-supervised manner, taking into account the teacher uncertainty. Given the vast amount of unannotated data and the challenges associated with capturing annotated depth in medical minimally invasive procedures, we advocate a fully self-supervised approach that only requires RGB images and the geometric and photometric calibration of the endoscope. In endoscopic imaging, the camera and light sources are co-located at a small distance from the target surfaces. This setup indicates that brighter areas of the image are nearer to the camera, while darker areas are further away. Building on this observation, we exploit the fact that for any given albedo and surface orientation, pixel brightness is inversely proportional to the square of the distance. We propose the use of illumination as a strong single-view self-supervisory signal for deep neural networks.

Bayesian Deep Neural Networks for Supervised Learning of Single-View Depth

Uncertainty and Self-Supervision in Single-View Depth

On the Uncertain Single-View Depths in Colonoscopies

Bayesian DeNet: Monocular Depth Prediction and Frame-Wise Fusion with Synchronized Uncertainty

Depth Estimation from Monocular Images Using Dilated Convolution and Uncertainty Learning.

Evaluation of Multi-task Uncertainties in Joint Semantic Segmentation and Monocular Depth Estimation

Parameter-efficient Bayesian Neural Networks for Uncertainty-aware Depth Estimation

Uncertainty-Aware Self-Improving Framework for Depth Estimation

Self-Supervised Monocular Depth Estimation by Digging into Uncertainty Quantification

Uncertainty estimation in Deep Learning for Panoptic segmentation

MonoProb: Self-Supervised Monocular Depth Estimation with Interpretable Uncertainty

Self-Supervised Monocular Depth Estimation Via Discrete Strategy and Uncertainty

Measuring and Modeling Uncertainty Degree for Monocular Depth Estimation

Leveraging the Bhattacharyya coefficient for uncertainty quantification in deep neural networks

RBUE: A ReLU-Based Uncertainty Estimation Method of Deep Neural Networks

Bayesian deep learning of affordances from RGB images

Self-supervised Monocular Depth Estimation with Uncertainty-aware Feature Enhancement and Depth Fusion

On Monocular Depth Estimation and Uncertainty Quantification using Classification Approaches for Regression

Uncertainty Quantification in Depth Estimation via Constrained Ordinal Regression

What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?

Deep Modeling of Non-Gaussian Aleatoric Uncertainty