Please use this identifier to cite or link to this item:
Publication type: Conference paper
Type of review: No review
Title: Efficient deep CNNs for cross-modal automated computer vision under time and space constraints
Authors: Amirian, Mohammadreza
Rombach, Katharina
Tuggener, Lukas
Schilling, Frank-Peter
Stadelmann, Thilo
et. al: No
DOI: 10.21256/zhaw-18357
Conference details: ECML-PKDD 2019, Würzburg, Germany, 16-19 September 2019
Issue Date: 2019
Publisher / Ed. Institution: ZHAW Zürcher Hochschule für Angewandte Wissenschaften
Language: English
Subjects: AutoDL; Automated deep learning; Convolutional neural networks; MobileNetV2; EfficientNet
Subject (DDC): 006: Special computer methods
Abstract: We present an automated computer vision architecture to handle video and image data using the same backbone networks. We show empirical results that lead us to adopt MOBILENETV2 as this backbone architecture. The paper demonstrates that neural architectures are transferable from images to videos through suitable preprocessing and temporal information fusion.
Fulltext version: Accepted version
License (according to publishing contract): Licence according to publishing contract
Departement: School of Engineering
Organisational Unit: Institute of Applied Information Technology (InIT)
Published as part of the ZHAW project: Ada – Advanced Algorithms for an Artificial Data Analyst
Appears in collections:Publikationen School of Engineering

Files in This Item:
File Description SizeFormat 
2019_Stadelmann_Efficient Deep CNNS_ECML_PKDD_2019.pdf168.59 kBAdobe PDFThumbnail

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.