Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) 2018, pp.1-3
Language
English
Type
Conference Paper
Abstract
Recently, deep architectures for depth estimation have been proposed for many applications as depth is a very essential element of 3D geometry in computer vision. However, we still need to make the architectures efficient in terms of model size and computation complexity in order for them to be adopted in mobile environments. While model size has been reduced greatly current state-of-the-art compression techniques, the amount of computation has not been much. In this paper, we propose a lightweight architecture that also reduces computational complexity by adopting RR(Reduction and Reconstruction) block. The design principles are motivated by Tiny-Darknet which is one of image classification architecture. We basically refer to existing compression techniques, but limit maximum number of convolutional layer expansions that have been implemented for performance preservation. Through this, total amount of computation does not increase more than a certain amount. Our architecture shows significant architectural savings 30 ∼ 59× in the number of trainable parameters compared to existing architectures. Furthermore, we demonstrated our architecture on the mobile GPU hardware and affords from 3.88 ∼ 5.37× less energy consumption and 2.54 ∼ 3.11× faster runtime while allows only 1.66%p ∼ 2.08%p performance degradation compare to our baseline application.
KSP Keywords
3D geometry, Compression techniques, Computational complexity, Computer Vision(CV), Current state, Deep architecture, Depth estimation, Essential element, GPU hardware, Image Classification, Light-weight
Copyright Policy
ETRI KSP Copyright Policy
The materials provided on this website are subject to copyrights owned by ETRI and protected by the Copyright Act. Any reproduction, modification, or distribution, in whole or in part, requires the prior explicit approval of ETRI. However, under Article 24.2 of the Copyright Act, the materials may be freely used provided the user complies with the following terms:
The materials to be used must have attached a Korea Open Government License (KOGL) Type 4 symbol, which is similar to CC-BY-NC-ND (Creative Commons Attribution Non-Commercial No Derivatives License). Users are free to use the materials only for non-commercial purposes, provided that original works are properly cited and that no alterations, modifications, or changes to such works is made. This website may contain materials for which ETRI does not hold full copyright or for which ETRI shares copyright in conjunction with other third parties. Without explicit permission, any use of such materials without KOGL indication is strictly prohibited and will constitute an infringement of the copyright of ETRI or of the relevant copyright holders.
J. Kim et. al, "Trends in Lightweight Kernel for Many core Based High-Performance Computing", Electronics and Telecommunications Trends. Vol. 32, No. 4, 2017, KOGL Type 4: Source Indication + Commercial Use Prohibition + Change Prohibition
J. Sim et.al, “the Fourth Industrial Revolution and ICT – IDX Strategy for leading the Fourth Industrial Revolution”, ETRI Insight, 2017, KOGL Type 4: Source Indication + Commercial Use Prohibition + Change Prohibition
If you have any questions or concerns about these terms of use, or if you would like to request permission to use any material on this website, please feel free to contact us
KOGL Type 4:(Source Indication + Commercial Use Prohibition+Change Prohibition)
Contact ETRI, Research Information Service Section
Privacy Policy
ETRI KSP Privacy Policy
ETRI does not collect personal information from external users who access our Knowledge Sharing Platform (KSP). Unathorized automated collection of researcher information from our platform without ETRI's consent is strictly prohibited.
[Researcher Information Disclosure] ETRI publicly shares specific researcher information related to research outcomes, including the researcher's name, department, work email, and work phone number.
※ ETRI does not share employee photographs with external users without the explicit consent of the researcher. If a researcher provides consent, their photograph may be displayed on the KSP.