ICML 2020 Workshop on

Uncertainty & Robustness in Deep Learning

July 17, 2020

Note: The UDL workshop will take place virtually this year.

Overview

There is growing interest in making deep neural networks more reliable. Challenges arise when models receive samples drawn from outside the training distribution. For example, a neural network tasked with classifying handwritten digits may assign high confidence predictions to cat images. Anomalies are frequently encountered when deploying ML models in the real world. Well-calibrated predictive uncertainty estimates are indispensable for many machine learning applications, such as self-driving vehicles and medical diagnosis systems. Generalization to unforeseen and worst-case inputs is also essential for robustness to distributional shift. In order to have ML models reliably predict in open environments, we must deepen technical understanding in the emerging areas of:

Learning algorithms that can detect changes in data distribution (e.g., out-of-distribution examples).
Mechanisms to estimate and calibrate confidence produced by neural networks in typical and unforeseen scenarios.
Methods to improve out-of-distribution generalization, including generalization to temporal, geographical, hardware, adversarial, stylization, and image-quality changes.
Benchmark datasets and protocols for evaluating model performance under distribution shift.
Key applications of robust and uncertainty-aware deep learning (e.g., computer vision, robotics, self-driving vehicles, medical imaging) as well as broader machine learning tasks.

This workshop will bring together researchers and practitioners from the machine learning communities, and highlight recent work that contributes to addressing these challenges. Our agenda will feature contributed papers with invited speakers. Through the workshop we hope to help identify fundamentally important directions on robust and reliable deep learning, and foster future collaborations.

Topics of interest include but not limited to:

• Model uncertainty estimation and calibration

• Probabilistic (Bayesian and non-Bayesian) neural networks

• Robustness to distribution shift and out-of-distribution generalization

• Anomaly detection and out-of-distribution detection

• Model misspecification

• Quantifying different types of uncertainty (known unknowns, unknown unknowns, contextual anomalies, ambiguities)

• Open world recognition and open set learning

• Connections between out-of-distribution generalization and adversarial robustness

• New datasets and protocols for evaluating uncertainty and robustness

Please see the call for papers for formatting instructions and deadlines.