A Study on Different Evolution in Computer Vision
Keywords:
Computer Vision(CV), Convolution, Convolution Neural network(CNN), Gradient, improvement in CV after CNN, Machine Learning, possible improvement in CVAbstract
Computer vision, when a computer and/or machine have sight, can be used in many applications like OCR, Vision Biometrics, Object Recognition, Social Media, Smart Cars etc., Different approach evolved over a period of time in computer vision problems, which can be categorized as, one after the deep learning in computer vision problem and the other before deep learning in computer vision problem. The prior one named as classical approach (HOG & SIFT., etc), could not learn from discrimination features from images and non adoptive for diverse image and doesn’t meet human level of accuracy. So there arises a requirement for learning method in computer vision Problems. Machine learning gives computers the ability to learn without being explicitly programmed. Deep learning or machine learning overcomes the drawbacks of classical approach by learning the features in the images and the diversity in the images implicitly and thus meets more accuracy than human vision. In this paper we will study difference methods like Classical & Deep learning for image classification problems , and analyze the draw backs and how the other approach overcome the drawbacks and accuracy levels meet by these approaches over the years
References
[1] Fei Fei Li Ted Talk: “How we teach computers to understand pictures starring”, establisher of ImageNet database.
[2]Erik G. & Learned-Miller "Introduction to Computer Vision" De-partment of Computer Science University of Massachusetts, Am-herst.
[3]David A. Forsyth, Jean Ponce "Computer Vision: A Modern Approach".
[4]Andrew ng Coursera Convolution Neural Net-works https://www.coursera.org/learn/convolutional-neural-networks
[5] "Three –Dimensional Computer Vision A Geometric viewpoint" @ 1993 Mas-sachusetts Institute of Technology
[6]Mubarak Shah "Computer vision Presentations"
[7]D.A Forsyth "Edge, Orientation ,HOGand Sift" http://luthuli.cs.uiuc.edu/~daf/courses/ComputerVisionTutorial2012/EdgesOrientationHOGSIFT-2012.pdf
[8]Y. Bengio; A. Courville; P. Vincent "Representation Learn-ing: A Review and New Perspectives". (2013) IEEE Trans. PAMI, special issue Learning Deep Architectures. 35: 1798–1828. ar-Xiv:1206.5538. doi:10.1109/tpami.2013.50
[9]Deep learning http://www.deeplearningbook.org/version-2016-01-14/contents/intro.html
[10]John McCarthy & Edward Feigenbaum . In Memoriam Arthur Samuel: "pioneer in machine learning"., 1990, AI Magazine. AAAI. 11 (3)
[11]Phil Simon "Too big to ignore: the business case for big data", 2013, Wiley. p. 89. ISBN 978-1-118-63817-0.
[12] Andrew NG courser – Deep lear-ninghttps://www.coursera.org/specializations/deep-learning
[13]Denny Atkin. "Computer Shopper: The Right GPU for You". Arc-hived from the original on 2007-05-06. Retrieved 2007-05-15.
[14]"Deep Learning Framework" ----https://developer.nvidia.com/deep-learning-frameworks
[15]Power of IOT & Big Data https://www.zdnet.com/topic/the-power-of-iot-and-big-data/
[16]Hamed Habibi Agh-dam; Elnaz Jahani Heravi "Guide to Convolutional Neural Networks A Practical Application to Traffic-Sign Detection and Classification" , Springer International Publishing AG
[17]LeCun et al., "Gradient-based learning to document recogni-tion", 1988
[18]Krizhevasky et al., "ImageNet classification with deep convo-lutional neural network" ,2012
[19]Simonyan & Zisserman , " Very deep convolutional networks for large-scale image recognition", 2015
[20]Chirstian Szegedy et al., 2015
[21]He et al., "Deep residual networks for image recognition" , 2015
[22]R. Dillmann et al., “Analysis and Optimization of Convolu-tional Neural Network Architectures” 2017
[24]http://mccormickml.com/2013/05/09/hog-person-detector-tutorial/
[25]https://en.wikipedia.org/wiki/Feature_learning
[26]http://mediatum.ub.tum.de/doc/1320382/document.pdf
[28]https://www.scribd.com/document/238827907/Opencv-Python-Tutroals
[29]https://en.wikipedia.org/wiki/Medical_imaging
[30]http://slazebni.cs.illinois.edu/spring17/lec01_cnn_architectures.pdf
[31]https://mesin-belajar.blogspot.com/2017/08/must-read-path-breaking-papers-about.html
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors contributing to this journal agree to publish their articles under the Creative Commons Attribution 4.0 International License, allowing third parties to share their work (copy, distribute, transmit) and to adapt it, under the condition that the authors are given credit and that in the event of reuse or distribution, the terms of this license are made clear.
