• Corpus ID: 246706224

Image-to-Image Regression with Distribution-Free Uncertainty Quantification and Applications in Imaging

  title={Image-to-Image Regression with Distribution-Free Uncertainty Quantification and Applications in Imaging},
  author={Anastasios Nikolas Angelopoulos and Amit Kohli and Stephen Bates and Michael I. Jordan and Jitendra Malik and Thayer Alshaabi and Srigokul Upadhyayula and Yaniv Romano},
Image-to-image regression is an important learning task, used frequently in biological imaging. Current algorithms, however, do not generally offer statistical guarantees that protect against a model’s mistakes and hallucinations. To ad-dress this, we develop uncertainty quantification techniques with rigorous statistical guarantees for image-to-image regression problems. In particular, we show how to derive uncertainty intervals around each pixel that are guaranteed to contain the true value… 

Figures from this paper

A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification

This hands-on introduction is aimed at a reader interested in the practical implementation of distribution-free UQ who is not necessarily a statistician, allowing them to rigorously quantify algorithmic uncertainty with one self-contained document.

Integrative conformal p-values for powerful out-of-distribution testing with labeled outliers

Novel conformal methods to test whether a new observation was sampled from the same distribution as a reference set are developed and are shown to outperform standard conformal pvalues through simulations as well as applications to image recognition and tabular data.

Training Uncertainty-Aware Classifiers with Conformalized Deep Learning

A novel training algorithm is developed that can lead to smaller conformal prediction sets with higher conditional coverage, after exact calibration with hold-out data, compared to state-of-the-art alternatives.

Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control

The main insight is to reframe the risk-control problem as multiple hypothesis testing, enabling techniques and mathematical arguments to be reframe from those in the previous literature.

Risk Control for Online Learning Models

This work develops a framework for constructing uncertainty sets that provably control risk—such as coverage of confidence intervals, false negative rate, or F1 score—in the online setting and extends conformal prediction to apply to a larger class of online learning problems.

Semantic uncertainty intervals for disentangled latent spaces

This work provides principled uncertainty intervals that are guaranteed to contain the true semantic factors for any underlying generative model, and reliably communicates semantically meaningful, principled, and instance-adaptive uncertainty in inverse problems like image super-resolution and image completion.

Improving Trustworthiness of AI Disease Severity Rating in Medical Imaging with Ordinal Conformal Prediction Sets

New uncertainty quantification methods are applied to a deep-learning model for grading the severity of spinal stenosis in lumbar spine MRI that provide tight coverage with small prediction set sizes and demonstrate a technique for forming ordinal prediction sets that are guaranteed to contain the correct stenosis severity within a user-deflned probability (conffdence interval).

Conformal Methods for Quantifying Uncertainty in Spatiotemporal Data: A Survey

Recent works on uncertainty quantification (UQ) for deep learning are surveyed, in particular distribution-free Conformal Prediction method for its mathematical properties and wide applicability and the role of UQ in the context of safe decision making is discussed.

Predictive Inference with Feature Conformal Prediction

This paper proposes feature conformal prediction, which extends the scope of conformal predictions to semantic feature spaces by leveraging the inductive bias of deep representation learning and demonstrates that feature conformAL prediction provably outperforms regular conformal Prediction under mild assumptions.

Conformal Prediction with Temporal Quantile Adjustments

Temporal Quantile Adjustment (TQA) is developed, a general method to construct efficient and valid prediction intervals (PIs) for regression on cross-sectional time series data and validate TQA’s performance through extensive experimentation.



Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network

SRGAN, a generative adversarial network (GAN) for image super-resolution (SR), is presented, to its knowledge, the first framework capable of inferring photo-realistic natural images for 4x upscaling factors and a perceptual loss function which consists of an adversarial loss and a content loss.

Content-aware image restoration: pushing the limits of fluorescence microscopy

This work shows how content-aware image restoration based on deep learning extends the range of biological phenomena observable by microscopy by bypassing the trade-offs between imaging speed, resolution, and maximal light exposure that limit fluorescence imaging to enable discovery.

A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification

This hands-on introduction is aimed at a reader interested in the practical implementation of distribution-free UQ who is not necessarily a statistician, allowing them to rigorously quantify algorithmic uncertainty with one self-contained document.

Interval Neural Networks: Uncertainty Scores

A fast, non-Bayesian method for producing uncertainty scores in the output of pre-trained deep neural networks (DNNs) using a data-driven interval propagating network that produces fast, theoretically justified uncertainty scores for DNNs that are easy to interpret, come with added information and pose as improved error proxies.

Image-to-Image Translation with Conditional Adversarial Networks

Conditional adversarial networks are investigated as a general-purpose solution to image-to-image translation problems and it is demonstrated that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks.

U-Net: Convolutional Networks for Biomedical Image Segmentation

It is shown that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks.

Uncertainty Estimates and Multi-hypotheses Networks for Optical Flow

A new network architecture and loss function is introduced that enforce complementary hypotheses and provide uncertainty estimates efficiently with a single forward pass and without the need for sampling or ensembles.

Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning

A new theoretical framework is developed casting dropout training in deep neural networks (NNs) as approximate Bayesian inference in deep Gaussian processes, which mitigates the problem of representing uncertainty in deep learning without sacrificing either computational complexity or test accuracy.

Perceptual Losses for Real-Time Style Transfer and Super-Resolution

This work considers image transformation problems, and proposes the use of perceptual loss functions for training feed-forward networks for image transformation tasks, and shows results on image style transfer, where aFeed-forward network is trained to solve the optimization problem proposed by Gatys et al. in real-time.

Image denoising review: From classical to state-of-the-art approaches