Image Classification Using Modified DCNNs Trained by Cascaded Softmax and Generalized Large - Margin loss
Author(s):
Komma Vinuthna
ISSN:
2349-6002
Cite This Article:
Image Classification Using Modified DCNNs Trained by Cascaded Softmax and Generalized Large - Margin lossInternational Journal of Innovative Research in Technology(www.ijirt.org) ,ISSN: 2349-6002 ,Volume 6 ,Issue 6 ,Page(s):69-71 ,November 2019 ,Available :IJIRT148768_PAPER.pdf
Keywords:
Cascaded softmax loss, deep convolutional neural network (DCNN), fine-grained image classification, generalized large-margin (GLM) loss, hierarchical label structure
Abstract
I have developed a fine-grained image classifier using a general deep convolutional neural network (DCNN). I improved the fine-grained image classification accuracy of a DCNN model from the following two aspects. First, to better model the h- level hierarchical label structure of the fine-grained image classes contained in the given training data set, i introduced h fully connected (fc) layers to replace the top fc layer of a given DCNN model and train them with the cascaded softmax loss. Second, i proposed a novel loss function, namely, generalized large-margin (GLM) loss, to make the given DCNN model explicitly explore the hierarchical label structure and the similarity regularities of the fine-grained image classes. The GLM loss explicitly not only reduces between-class similarity and within-class variance of the learned features by DCNN models but also makes the subclasses belonging to the same coarse class be more similar to each other than those belonging to different coarse classes in the feature space. Moreover, the proposed fine-grained image classification framework is independent and can be applied to any DCNN structures. Comprehensive experimental evaluations of several general DCNN models (AlexNet, GoogLeNet, and VGG) using three benchmark data sets (Stanford car, fine-grained visual classification-aircraft, and CUB-200-2011) for the fine-grained image classification task demonstrate the effectiveness of my method.
Article Details
Unique Paper ID: 148768

Publication Volume & Issue: Volume 6, Issue 6

Page(s): 69 - 71
Article Preview & Download




Go To Issue



Call For Paper

Volume 6 Issue 4

Last Date 25 September 2019


About Us

IJIRT.org enables door in research by providing high quality research articles in open access market.

Send us any query related to your research on editor@ijirt.org

Social Media

Google Verified Reviews

Contact Details

Telephone:8200 61 5067
Email: editor@ijirt.org
Website: ijirt.org

Policies