Teacher-Free Knowledge Distillation Based on Non-Progressive Meta-Learned Simulated Annealing

Pin Hsuan Ho, Bing Ru Jiang, Albert S. Lin

研究成果: Conference contribution同行評審

1 引文 斯高帕斯(Scopus)

摘要

Knowledge distillation (KD) is one of the effective methods in model compression. Recently, some teacher-free knowledge distillation (TFKD) methods have been proposed to overcome the problem that the teacher model is too huge in the traditional KD. In this paper, we proposed a TFKD framework based on a non-progressive method that the target distribution will change during training. We use simulated annealing (SA) to control the variety of free-form distribution. In addition, we also developed a meta-learned reinforcement learning (RL) method to control the energy calculation in SA. We conducted the experiments on CIFARIOO, CIFARIO, and SVHN with the VGG-8 model. The accuracy of the VGG-8 model can be improved to 72.88%, which is higher than the baseline model by 2.44%. Compared to another teacher-free method currently, the performance of our method is also the highest.

原文English
主出版物標題2024 IEEE 3rd International Conference on Computing and Machine Intelligence, ICMI 2024 - Proceedings
編輯Ahmed Abdelgawad, Akhtar Jamil, Alaa Ali Hameed
發行者Institute of Electrical and Electronics Engineers Inc.
ISBN(電子)9798350372977
DOIs
出版狀態Published - 2024
事件3rd IEEE International Conference on Computing and Machine Intelligence, ICMI 2024 - Mt. Pleasant, 美國
持續時間: 13 4月 202414 4月 2024

出版系列

名字2024 IEEE 3rd International Conference on Computing and Machine Intelligence, ICMI 2024 - Proceedings

Conference

Conference3rd IEEE International Conference on Computing and Machine Intelligence, ICMI 2024
國家/地區美國
城市Mt. Pleasant
期間13/04/2414/04/24

指紋

深入研究「Teacher-Free Knowledge Distillation Based on Non-Progressive Meta-Learned Simulated Annealing」主題。共同形成了獨特的指紋。

引用此