E-Stitchup: Data Augmentation for Pre-Trained Embeddings
원문 초록
In this work, we propose data augmentation methods for embeddings from pre-trained deep learning models that take a weighted combination of a pair of input embeddings, as inspired by Mixup, and combine such augmentation with extra label softening. These methods are shown to significantly increase classification accuracy, reduce training time, and improve confidence calibration of a downstream model that is trained with them. As a result of such improved confidence calibration, the model output can be more intuitively interpreted and used to accurately identify out-of-distribution data by applying an appropriate confidence threshold to model predictions. The identified out-of-distribution data can then be prioritized for labeling, thus focusing labeling effort on data that is more likely to boost model performance. These findings, we believe, lay a solid foundation for improving the classification performance and calibration of models that use pre-trained embeddings as input and provide several benefits that prove extremely useful in a production-level deep learning system.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
PSAT: Pediatric Segmentation Approaches via Adult Augmentations and Transfer Learning
DS@GT at CheckThat! 2025: Detecting Subjectivity via Transfer-Learning and Corrective Data Augmentation
SSS: Semi-Supervised SAM-2 with Efficient Prompting for Medical Imaging Segmentation
Data-Efficient Challenges in Visual Inductive Priors: A Retrospective
An Explainable Deep Learning Framework for Brain Stroke and Tumor Progression via MRI Interpretation
Facial Beauty Analysis Using Distribution Prediction and CNN Ensembles
토론
이 논문과 연결된 포럼 글이 아직 없습니다.