Safety

Two Coupled Rejection Metrics Can Tell Adversarial Examples Apart CVPR 2022 Anti-Backdoor Learning: Training Clean Models on Poisoned Data NeurIPS 2021 Conservative Safety Critics for Exploration ICLR 2021 Counterexample Guided RL Policy Refinement Using Bayesian Optimization NeurIPS 2021 FaceSec: A Fine-Grained Robustness Evaluation Framework for Face Recognition Systems CVPR 2021 Fooling Thermal Infrared Pedestrian Detectors in Real World Using Small Bulbs AAAI 2021 Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violations NeurIPS 2021 Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs NeurIPS 2021 Leveraging on Deep Reinforcement Learning for Autonomous Safe Decision-Making in Highway On-ramp Merging (Student Abstract) AAAI 2021 Mind the Style of Text! Adversarial and Backdoor Attacks Based on Text Style Transfer EMNLP 2021 Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database AAAI 2021 Probabilistic Robustness Quantification of Neural Networks AAAI 2021 Profanity-Avoiding Training Framework for Seq2seq Models with Certified Robustness EMNLP 2021 REM: Efficient Semi-Automated Real-Time Moderation of Online Forums ACL 2021 Safe Policy Optimization with Local Generalized Linear Function Approximations NeurIPS 2021 Safe Reinforcement Learning by Imagining the Near Future NeurIPS 2021 Safe Reinforcement Learning with Linear Function Approximation ICML 2021 Safety Assurance for Systems with Machine Learning Components AAAI 2021 Sample-Specific Output Constraints for Neural Networks AAAI 2021 Tightening Robustness Verification of Convolutional Neural Networks with Fine-Grained Linear Approximation AAAI 2021 Topological Detection of Trojaned Neural Networks NeurIPS 2021 Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds NeurIPS 2021 Transfer Learning for Efficient Iterative Safety Validation AAAI 2021 Abstract Constraints for Safe and Robust Robot Learning from Demonstration AAAI 2020 Certifying Confidence via Randomized Smoothing NeurIPS 2020