Learning to Annotate Delayed and False AEB Events: A Practical System for Extreme Class Imbalance and Asymmetric Label Noise
Mengxiang Hao, Xin Jiang, Xinghao Huang, Wenliang Su, Zhiteng Wang, Junjie Rao, Xiaotian Yang, Wei Liao, Chengyu Han, Gen Liang, Yulun Song, Zhitao Xu, Xianpeng Lang
THE PROBLEM
This paper focuses on learning. Proposes automated Data, Distributions & Training IssuesAnnotationHuman-provided labels or metadata attached to data. system for Autonomous Emergency Braking edge cases using Data, Distributions & Training IssuesData augmentationArtificially varying training data to improve generalization. (focal attribute Manipulation & TasksManipulationUsing a robot arm or hand to move or interact with objects., ego-dynamics transplanting, agent masking) and Data, Distributions & Training IssuesNoiseUnwanted variation or randomness in sensor readings or actuation. suppression (stable hardness estimation, probe-guided thresholding) to handle extreme class imbalance where delayed/false triggers <5% of samples. Read the paper by tracking the Core ConceptsTaskThe job the robot is supposed to complete, such as pick-and-place, navigation, or drawer opening. definition, the Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. or data assumptions, and the evidence that supports the claimed improvement.
HOW IT WORKS
Task framing
Core method
Data and supervision
Evaluation evidence
KEY RESULTS
This paper solves a production robotics Data, Distributions & Training IssuesAnnotationHuman-provided labels or metadata attached to data. bottleneck: automatically labeling rare critical AEB failures (false/delayed triggers) from thousands of daily events, achieving 80% higher recall on edge cases while cutting manual labeling work by 50%. The framework handles extreme class imbalance and mislabeled majority samples—two problems that kill minority-class learning—using Simulation & Sim-to-RealSynthetic dataArtificially generated training data, often from simulation. augmentation and noise-robust Robot LearningTrainingThe process of fitting a model using data or experience..
WHY DEVELOPERS SHOULD CARE
This paper solves a production robotics Data, Distributions & Training IssuesAnnotationHuman-provided labels or metadata attached to data. bottleneck: automatically labeling rare critical AEB failures (false/delayed triggers) from thousands of daily events, achieving 80% higher recall on edge cases while cutting manual labeling work by 50%. The framework handles extreme class imbalance and mislabeled majority samples—two problems that kill minority-class learning—using Simulation & Sim-to-RealSynthetic dataArtificially generated training data, often from simulation. augmentation and noise-robust Robot LearningTrainingThe process of fitting a model using data or experience..
LIMITATIONS
The main limitation to check is whether the claimed behavior holds outside the paper's reported setup. That means testing across different Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. embodiments, scenes, objects, and data distributions.
WHAT COMES NEXT
The practical next step is independent reproduction with clear baselines, ablations, and stress tests. For a developer, the useful follow-up is to map the paper's learning assumptions onto a concrete Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. stack, then test the smallest version of the method that could run end to end.