Positive-Congruent Training: Towards Regression-Free Model Updates
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021(2021)
摘要
Reducing inconsistencies in the behavior of different versions of an AI system can be as important in practice as reducing its overall error. In image classification, sample-wise inconsistencies appear as "negative flips": A new model incorrectly predicts the output for a test sample that was correctly classified by the old (reference) model. Positive-congruent (PC) training aims at reducing error rate while at the same time reducing negative flips, thus maximizing congruency with the reference model only on positive predictions, unlike model distillation. We propose a simple approach for PC training, Focal Distillation, which enforces congruence with the reference model by giving more weights to samples that were correctly classified. We also found that, if the reference model itself can be chosen as an ensemble of multiple deep neural networks, negative flips can be further reduced without affecting the new model's accuracy.
更多查看译文
关键词
regression-free model updates,AI system,image classification,sample-wise inconsistencies,negative flips,positive-congruent training,error rate reduction,congruency,positive predictions,model distillation,PC training,congruence,focal distillation,deep neural network
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络