每年專案
摘要
Convolutional Neural Networks (CNNs) are widely applied in various computer version applications such as object recognition and image classification. With large amount of Multiply Accumulate Operations (MACs) in CNN computations, it is a trend to process the data locally at edge devices to avoid significant data movement. MobileNets were proposed as a light-weight neural network for mobile and embedded devices by using depthwise separable convolutions. Moreover, to further reduce commutation efforts, quantization can be appropriately applied to MobileNets without significant accuracy loss. To provide an efficient computation platform for quantized MobileNets, in this paper we propose a novel performance-Aware reconfigurable accelerator design which prefect fit the depthwise separable convolutions. Experimental results show that we can achieve 16% area reduction and 1.07x speedup compared to previous MobileNets accelerator design.
原文 | ???core.languages.en_GB??? |
---|---|
主出版物標題 | Proceedings - International SoC Design Conference 2021, ISOCC 2021 |
發行者 | Institute of Electrical and Electronics Engineers Inc. |
頁面 | 290-291 |
頁數 | 2 |
ISBN(電子) | 9781665401746 |
DOIs | |
出版狀態 | 已出版 - 2021 |
事件 | 18th International System-on-Chip Design Conference, ISOCC 2021 - Jeju Island, Korea, Republic of 持續時間: 6 10月 2021 → 9 10月 2021 |
出版系列
名字 | Proceedings - International SoC Design Conference 2021, ISOCC 2021 |
---|
???event.eventtypes.event.conference???
???event.eventtypes.event.conference??? | 18th International System-on-Chip Design Conference, ISOCC 2021 |
---|---|
國家/地區 | Korea, Republic of |
城市 | Jeju Island |
期間 | 6/10/21 → 9/10/21 |