p-Meta: Towards on-device deep model adaptation

• Main Authors: QU, Zhongnan, ZHOU, Zimu, TONG, Yongxin, THIELE, Lothar
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: deep neural networks; memory-efficient training; meta learning; Numerical Analysis and Scientific Computing
• Online Access: https://ink.library.smu.edu.sg/sis_research/7275; https://ink.library.smu.edu.sg/context/sis_research/article/8278/viewcontent/3534678.3539293_pvoa.pdf
• Institution: Singapore Management University

Data collected by IoT devices are often private and have a large diversity across users. Therefore, learning requires pre-training a model with available representative data samples, deploying the pre-trained model on IoT devices, and adapting the deployed model on the device with local data. Such an on-device adaption for deep learning empowered applications demands data and memory efficiency. However, existing gradient-based meta learning schemes fail to support memory-efficient adaptation. To this end, we propose p-Meta, a new meta learning method that enforces structure-wise partial parameter updates while ensuring fast generalization to unseen tasks. Evaluations on few-shot image classification and reinforcement learning tasks show that p-Meta not only improves the accuracy but also substantially reduces the peak dynamic memory by a factor of 2.5 on average compared to state-of-the-art few-shot adaptation methods.