SafeAI 2021
Artificial Intelligence Safety 2021

Table of Contents

• Preface
  
  

Session 1: Dynamic Safety and Anomaly Assessment

• Feature Space Singularity for Out-of-Distribution Detection
  Haiwen Huang, Zhihan Li, Lulu Wang, Sishuo Chen, Xinyu Zhou, Bin Dong
• An Evaluation of “Crash Prediction Networks” (CPN) for Autonomous Driving Scenarios in CARLA Simulator
  Saasha Nair, Sina Shafaei, Daniel Auge, Alois Knoll
• From Black-box to White-box: Examining Confidence Calibration under different Conditions
  Franziska Schwaiger, Maximilian Henne, Fabian Küppers, Felippe Schmoeller Roza, Karsten Roscher, Anselm Haselhoff

Session 2: Safety Considerations for the Assurance of AI-based Systems

• The Utility of Neural Network Test Coverage Measures
  Rob Ashmore, Alec Banks
• Safety Properties of Inductive Logic Programming
  Gavin Leech, Nandi Schoots, Joar Skalse
• A Hybrid-AI Approach for Competence Assessment of Automated Driving functions
  Jan-Pieter Paardekooper, Mauro Comi, Corrado Grappiolo, Ron Snijders, Willeke van Vught, Rutger Beekelaar

Session 3: Adversarial Machine Learning and Trustworthiness

• Adversarial Robustness for Face Recognition: How to Introduce Ensemble Diversity among Feature Extractors?
  Takuma Amada, Kazuya Kakizaki, Seng Pei Liew, Toshinori Araki, Joseph Keshet, Jun Furukawa
• Multi-Modal Generative Adversarial Networks Make Realistic and Diverse but Untrustworthy Predictions When Applied to Ill-posed Problems
  John Hyatt, Michael Lee
• DeepFakesON-Phys: DeepFakes Detection based on Heart Rate Estimation
  Javier Hernandez-Ortega, Ruben Tolosana, Julian Fierrez, Aythami Morales

Session 4: Safe Autonomous Agents

• What criminal and civil law tells us about Safe RL techniques to generate law-abiding behaviour
  Hal Ashton
• Performance of Bounded-Rational Agents With the Ability to Self-Modify
  Jakub Tětek, Marek Sklenka, Tomáš Gavenčiak
• Deep CPT-RL: Imparting Human-Like Risk Sensitivity to Artificial Agents
  Jared Markowitz, Marie Chau, I-Jeng Wang
• Challenges for Using Impact Regularizers to Avoid Negative Side Effects
  David Lindner, Kyle Matoba, Alexander Meulemans

Poster Papers

• Towards an Ontological Framework for Environmental Survey Hazard Analysis of Autonomous Systems
  Christopher Harper, Praminda Caleb-Solly
• Overestimation Learning with Guarantees
  Adrien Gauffriau, François Malgouyres, Mélanie Ducoffe
• On the Use of Available Testing Methods for Verification & Validation of AI-based Software and Systems
  Franz Wotawa
• Runtime Decision Making Under Uncertainty in Autonomous Vehicles
  Vibhu Gautam, Youcef Gheraibia, Rob Alexander, Richard Hawkins
• Negative Side Effects and AI Agent Indicators: Experiments in SafeLife
  John Burden, Jose Hernandez-Orallo, Sean O'Heigeartaigh
• Time for AI (Ethics) Maturity Model Is Now
  Ville Vakkuri, Marianna Jantunen, Erika Halme, Kai-Kristian Kemell, Anh Nguyen-Duc, Tommi Mikkonen, Pekka Abrahamsson
• AI-Blueprint for Deep Neural Networks
  Ernest Wozniak, Henrik Putzer, Carmen Carlan
• Neural Criticality: Validation of Convolutional Neural Networks
  Vaclav Divis, Marek Hruz
• Adversarial Attacks for Tabular Data: Application to Fraud Detection and Imbalanced Data
  Francesco Cartella, Orlando Anunciacao, Yuki Funabiki, Daisuke Yamaguchi, Toru Akishita, Olivier Elshocht
• Correct-by-Construction Multi-Label Classification Networks
  Eleonora Giunchiglia, Thomas Lukasiewicz
• Classification Confidence Scores with Point-wise Guarantees
  Nivasini Ananthakrishnan, Shai Ben-David, Tosca Lechner