dc.contributor.author |
Mishra, Rahul |
|
dc.contributor.author |
Gupta, Hari Prabhat |
|
dc.date.accessioned |
2024-02-13T06:49:58Z |
|
dc.date.available |
2024-02-13T06:49:58Z |
|
dc.date.issued |
2023-07-19 |
|
dc.identifier.issn |
15361233 |
|
dc.identifier.uri |
http://localhost:8080/xmlui/handle/123456789/2891 |
|
dc.description |
This paper published with affiliation IIT (BHU), Varanasi in Open Access Mode. |
en_US |
dc.description.abstract |
Automated feature extraction capability and significant performance of Deep Neural Networks (DNN) make them suitable for Internet of Things (IoT) applications. However, deploying DNN on edge devices becomes prohibitive due to the colossal computation, energy, and storage requirements. This paper presents a novel approach, EarlyLight, for designing and training lightweight DNN using large-size DNN. The approach considers the available storage, processing speed, and maximum allowable processing time to execute the task on edge devices. We present a knowledge distillation based training procedure to train the lightweight DNN to achieve adequate accuracy. During the training of lightweight DNN, we introduce a novel early halting technique, which preserves network resources; thus, speedups the training procedure. Finally, we present the empirically and real-world evaluations to verify the effectiveness of the proposed approach under different constraints using various edge devices. |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
Institute of Electrical and Electronics Engineers Inc. |
en_US |
dc.relation.ispartofseries |
IEEE Transactions on Mobile Computing; |
|
dc.subject |
Artificial neural networks |
en_US |
dc.subject |
Deep neural networks |
en_US |
dc.subject |
Internet of Things |
en_US |
dc.subject |
knowledge distillation |
en_US |
dc.subject |
Knowledge engineering |
en_US |
dc.subject |
Mobile computing |
en_US |
dc.subject |
Performance evaluation |
en_US |
dc.subject |
Task analysis |
en_US |
dc.subject |
Training |
en_US |
dc.title |
Designing and Training of Lightweight Neural Networks on Edge Devices using Early Halting in Knowledge Distillation |
en_US |
dc.type |
Article |
en_US |