VALSE 首页 活动通知 好文作者面授招 查看内容

20170412-06 董宣毅:More is Less: A More Complicated Network with...

2017-4-5 10:18| 发布者: 程一-计算所| 查看: 6339| 评论: 0

摘要: 报告嘉宾2:董宣毅(University of Technology Sydeney)报告时间:2017年4月12日晚21:00(北京时间)报告题目:More is Less: A More Complicated Network with Less Inference Complexity主持人:郑良(University ...

报告嘉宾2:董宣毅(University of Technology Sydeney)


报告题目:More is Less: A More Complicated Network with Less Inference Complexity

主持人:郑良(University of Technology Sydeney)

报告摘要:In this paper, we present a novel and general network structure towards accelerating the inference process of convolutional neural networks, which is more complicated in network structure yet with less inference complexity. The core idea is to equip each original convolutional layer with another low-cost collaborative layer (LCCL), and the element-wise multiplication of the ReLU outputs of these two parallel layers produces the layer-wise output. The combined layer is potentially more discriminative than the original convolutional layer, and its inference is faster for two reasons: 1) the zero cells of the LCCL feature maps will remain zero after element-wise multiplication, and thus it is safe to skip the calculation of the corresponding high-cost convolution in the original convolutional layer; 2) LCCL is very fast if it is implemented as a 1*1 convolution or only a single filter shared by all channels. Extensive experiments on the CIFAR-10, CIFAR-100 and ILSCRC-2012 benchmarks show that our proposed network structure can accelerate the inference process by 32% on average with negligible performance drop.

文章信息:Xuanyi Dong, Junshi Huang, Yi Yang, Shuicheng Yan.  More is Less: A More Complicated Network with Less Inference Complexity. CVPR, 2017.

报告人简介:Xuanyi Dong is a first-year Ph.D student at University of Technology Sydeney (UTS), under the supervision of Associate Prof. Yi Yang. I received my B.E from School of Computer Science and Technology, Beihang University in 2016. My research interest is computer vision and deep learning techniques for video and image content understanding, such as object detection, 3D reconstruction, and video feature modelling.



小黑屋|手机版|Archiver|Vision And Learning SEminar

GMT+8, 2021-4-14 06:15 , Processed in 0.015523 second(s), 15 queries .

Powered by Discuz! X3.4

Copyright © 2001-2020, Tencent Cloud.