Emotional Analysis Using Image Processing
U.M Prakash1, Pratyush2, Pranshu Dixit3, Anamay Kumar Ojha4
1Mr. U.M Prakash, Department of Computer Science, SRM, IST (Delhi), India.
2Mr. Pratyush, Department of Computer Science SRM, IST (Delhi), India.
3Mr. Pranshu Dixit, Department of Computer Science SRM, IST (Delhi), India.
4Mr. Anamay Kumar Ojha, Department of Computer Science SRM, IST (Delhi), India.
Manuscript received on 06 February 2019 | Revised Manuscript received on 19 February 2019 | Manuscript Published on 04 March 2019 | PP: 258-262 | Volume-7 Issue-5S2 January 2019 | Retrieval Number: ES2043017519/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In machine learning, a convolutional neural network (CNN or Conv Net) is a part of deep and feed-forward artificial neural networks that has successfully visualized images. CNNs use a variation of multilayer perceptron designed to require minimal pre-processing. CNNs use relatively little preprocessing compared to other image classification algorithms. This means that the network learns the filtering that was hand engineered in other algorithms. This independence of human efforts for feature design is a major advantage due to which we are using it in our paper. In the context of machine vision, image recognition is the capability of software to identify people, places, objects, actions and writing in images. When we are using our algorithm train the model from our data set of around 600 images, we are getting an accuracy of 85.23%. We can also use other methods for modelling in for this problem set.
Keywords: Filter, Kernel Size, Convolving, Activation Map, Feature Map, Stride, Max Pool, Activation Function, Reception Field, Epoch Cycles.
Scope of the Article: Signal and Image Processing