On-Device Learning with Binary Neural Networks

被引:0
作者
Vorabbi, Lorenzo [1 ,2 ]
Maltoni, Davide [2 ]
Santi, Stefano [1 ]
机构
[1] Datal Labs, I-40012 Bologna, Italy
[2] Univ Bologna, DISI, Cesena Campus, I-47521 Cesena, Italy
来源
IMAGE ANALYSIS AND PROCESSING - ICIAP 2023 WORKSHOPS, PT I | 2024年 / 14365卷
关键词
Binary Neural Networks; On-device Learning; Continual Learning;
D O I
10.1007/978-3-031-51023-6_4
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Existing Continual Learning (CL) solutions only partially address the constraints on power, memory and computation of the deep learning models when deployed on low-power embedded CPUs. In this paper, we propose a CL solution that embraces the recent advancements in CL field and the efficiency of the Binary Neural Networks (BNN), that use 1-bit for weights and activations to efficiently execute deep learning models. We propose a hybrid quantization of CWR* (an effective CL approach) that considers differently forward and backward pass in order to retain more precision during gradient update step and at the same time minimizing the latency overhead. The choice of a binary network as backbone is essential to meet the constraints of low power devices and, to the best of authors' knowledge, this is the first attempt to prove on-device learning with BNN. The experimental validation carried out confirms the validity and the suitability of the proposed method.
引用
收藏
页码:39 / 50
页数:12
相关论文
共 24 条
[1]  
Banner R, 2018, Advances in neural information processing systems, V31
[2]  
Bannink T., 2021, Proc Mach Learn Syst, V3, P680
[3]  
Bethge J, 2019, Arxiv, DOI arXiv:1906.08637
[4]  
Cai H, 2019, Arxiv, DOI arXiv:1812.00332
[5]  
Das D, 2018, Arxiv, DOI [arXiv:1802.00930, DOI 10.48550/ARXIV.1802.00930]
[6]   A Continual Learning Survey: Defying Forgetting in Classification Tasks [J].
De Lange, Matthias ;
Aljundi, Rahaf ;
Masana, Marc ;
Parisot, Sarah ;
Jia, Xu ;
Leonardis, Ales ;
Slabaugh, Greg ;
Tuytelaars, Tinne .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (07) :3366-3385
[7]  
Howard AG, 2017, Arxiv, DOI arXiv:1704.04861
[8]  
Gupta S, 2015, PR MACH LEARN RES, V37, P1737
[9]  
Hubara I, 2016, ADV NEUR IN, V29
[10]  
Jacob B., 2017, GEMMLowp: A small self-contained lowprecision GEMM library