arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.07599cs.LG
收藏

一种基于模型压缩和迁移学习的对比知识转移框架

A Contrastive Knowledge Transfer Framework for Model Compression and Transfer Learning

Arizona State University(亚利桑那州立大学)

Kaiqi Zhao, Yitao Chen, Ming Zhao

摘要:

知识转移(KT)实现了有竞争力的性能,并广泛用于图像分类任务的模型压缩和迁移学习。现有的KT工程转移的信息从一个大的模型(“教师”)训练一个小的模型(“学生”),通过最小化的差异,他们的条件独立的输出分布。然而,这些工作忽略了从教师的中间表示,这导致有限的有效性,他们是由各种启发式的直觉,这使得它很难概括的高维结构知识。本文提出了一种新的对比知识转移框架(CKTF),它使足够的结构知识从教师转移到学生之间的中间表示,通过优化多个对比目标。此外,CKTF提供了一个普遍的协议,现有的KT技术,并通过将其作为CKTF的具体情况下,得到显着提高其性能。广泛的评估表明,CKTF在各种模型和数据集上的模型压缩和迁移学习方面始终优于现有KT作品0.04%至11.59%和0.4%至4.75%。

英文摘要:

Knowledge Transfer (KT) achieves competitive performance and is widely used for image classification tasks in model compression and transfer learning. Existing KT works transfer the information from a large model ("teacher") to train a small model ("student") by minimizing the difference of their conditionally independent output distributions. However, these works overlook the high-dimension structural knowledge from the intermediate representations of the teacher, which leads to limited effectiveness, and they are motivated by various heuristic intuitions, which makes it difficult to generalize. This paper proposes a novel Contrastive Knowledge Transfer Framework (CKTF), which enables the transfer of sufficient structural knowledge from the teacher to the student by optimizing multiple contrastive objectives across the intermediate representations between them. Also, CKTF provides a generalized agreement to existing KT techniques and increases their performance significantly by deriving them as specific cases of CKTF. The extensive evaluation shows that CKTF consistently outperforms the existing KT works by 0.04% to 11.59% in model compression and by 0.4% to 4.75% in transfer learning on various models and datasets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑