Object detection utilizing modified auto encoder and convolutional neural networks

Deep learning models are widely used in object detection area, including combination of multiple non-linear data transformations. The objective is receiving brief and concise information for feature representations. Due to the high volume of processing data, object detection in videos has been faced...

Full description

Bibliographic Details
Main Authors: Khiarak, Jalil Nourmohammadi, Mazaheri, Samaneh, Tayebi, Rohollah Moosavi, Noorbakhsh-Devlagh, Hamid
Format: Conference or Workshop Item
Language:English
Published: IEEE 2018
Online Access:http://psasir.upm.edu.my/id/eprint/69656/
http://psasir.upm.edu.my/id/eprint/69656/1/Object%20detection%20utilizing%20modified%20auto%20encoder%20and%20convolutional%20neural%20networks.pdf
_version_ 1848856469935489024
author Khiarak, Jalil Nourmohammadi
Mazaheri, Samaneh
Tayebi, Rohollah Moosavi
Noorbakhsh-Devlagh, Hamid
author_facet Khiarak, Jalil Nourmohammadi
Mazaheri, Samaneh
Tayebi, Rohollah Moosavi
Noorbakhsh-Devlagh, Hamid
author_sort Khiarak, Jalil Nourmohammadi
building UPM Institutional Repository
collection Online Access
description Deep learning models are widely used in object detection area, including combination of multiple non-linear data transformations. The objective is receiving brief and concise information for feature representations. Due to the high volume of processing data, object detection in videos has been faced with big challenges, such as mass calculation. To increase the object detection precision in videos, a hybrid method is proposed, in this paper. Some modifications are applied to auto encoder neural networks, for the compact and discriminative learning of object features. Furthermore, for object classification, firstly extracted features are transferred to a convolutional neural network, and after feature convolution with input pictures, they will be classified. The proposed method has two main advantages over other unsupervised feature learning techniques. Firstly, as it will be shown, features are detected with a much higher precision. Secondly, in the proposed method, the outcome is compact and additional unnecessary information is removed; while the existing unsupervised feature learning models mainly learn repeated and redundant information of the features. Experimental evaluation shows that precision of feature detection improved by 1.5% in average in compare with the state-of-the-art methods.
first_indexed 2025-11-15T11:42:10Z
format Conference or Workshop Item
id upm-69656
institution Universiti Putra Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T11:42:10Z
publishDate 2018
publisher IEEE
recordtype eprints
repository_type Digital Repository
spelling upm-696562019-07-08T02:44:00Z http://psasir.upm.edu.my/id/eprint/69656/ Object detection utilizing modified auto encoder and convolutional neural networks Khiarak, Jalil Nourmohammadi Mazaheri, Samaneh Tayebi, Rohollah Moosavi Noorbakhsh-Devlagh, Hamid Deep learning models are widely used in object detection area, including combination of multiple non-linear data transformations. The objective is receiving brief and concise information for feature representations. Due to the high volume of processing data, object detection in videos has been faced with big challenges, such as mass calculation. To increase the object detection precision in videos, a hybrid method is proposed, in this paper. Some modifications are applied to auto encoder neural networks, for the compact and discriminative learning of object features. Furthermore, for object classification, firstly extracted features are transferred to a convolutional neural network, and after feature convolution with input pictures, they will be classified. The proposed method has two main advantages over other unsupervised feature learning techniques. Firstly, as it will be shown, features are detected with a much higher precision. Secondly, in the proposed method, the outcome is compact and additional unnecessary information is removed; while the existing unsupervised feature learning models mainly learn repeated and redundant information of the features. Experimental evaluation shows that precision of feature detection improved by 1.5% in average in compare with the state-of-the-art methods. IEEE 2018 Conference or Workshop Item PeerReviewed text en http://psasir.upm.edu.my/id/eprint/69656/1/Object%20detection%20utilizing%20modified%20auto%20encoder%20and%20convolutional%20neural%20networks.pdf Khiarak, Jalil Nourmohammadi and Mazaheri, Samaneh and Tayebi, Rohollah Moosavi and Noorbakhsh-Devlagh, Hamid (2018) Object detection utilizing modified auto encoder and convolutional neural networks. In: 22nd IEEE Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA 2018), 19-21 Sept. 2018, Poznan, Poland. (pp. 43-49). 10.23919/SPA.2018.8563423
spellingShingle Khiarak, Jalil Nourmohammadi
Mazaheri, Samaneh
Tayebi, Rohollah Moosavi
Noorbakhsh-Devlagh, Hamid
Object detection utilizing modified auto encoder and convolutional neural networks
title Object detection utilizing modified auto encoder and convolutional neural networks
title_full Object detection utilizing modified auto encoder and convolutional neural networks
title_fullStr Object detection utilizing modified auto encoder and convolutional neural networks
title_full_unstemmed Object detection utilizing modified auto encoder and convolutional neural networks
title_short Object detection utilizing modified auto encoder and convolutional neural networks
title_sort object detection utilizing modified auto encoder and convolutional neural networks
url http://psasir.upm.edu.my/id/eprint/69656/
http://psasir.upm.edu.my/id/eprint/69656/
http://psasir.upm.edu.my/id/eprint/69656/1/Object%20detection%20utilizing%20modified%20auto%20encoder%20and%20convolutional%20neural%20networks.pdf