2020/06/06 by Deepan Das, Haley Massa, Das, Deepan +5 · 13 citations
Chemistry · Computer Science · Decision Sciences · #Artificial Intelligence (cs.AI) #Chemistry #Chromatography #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Data Quality and Management #Data Stream Mining Techniques #Data science #Distillation #Econometrics #Economics #FOS: Computer and information sciences #Machine Learning (cs.LG) #Privacy-Preserving Technologies in Data #cs.AI #cs.CV #cs.LG
paper · pdf · doi:10.48550/arxiv.2006.03810
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2020/06/06 · arxiv created 2020/06/09 · arxiv updated 2020/06/11 · openalex created_date 2020/06/12 · openalex updated_date 2026/07/28
Generalization Performance of Deep Learning models trained using Empirical Risk Minimization can be improved significantly by using Data Augmentation strategies such as simple transformations, or using Mixed Samples. We attempt to empirically analyze the impact of such strategies on the transfer of generalization between teacher and student models in a distillation setup. We observe that if a teacher is trained using any of the mixed sample augmentation strategies, such as MixUp or CutMix, the student model distilled from it is impaired in its generalization capabilities. We hypothesize that such strategies limit a model's capability to learn example-specific features, leading to a loss in quality of the supervision signal during distillation. We present a novel Class-Discrimination metric to quantitatively measure this dichotomy in performance and link it to the discriminative capacity induced by the different strategies on a network's latent space.