Custom Dataset Creation with Tensorflow Framework and Image Processing for Google T-Rex
Publication Date
9-25-2020
Document Type
Conference Proceeding
Publication Title
2020 12th International Conference on Computational Intelligence and Communication Networks (CICN)
DOI
10.1109/CICN49253.2020.9242565
First Page
45
Last Page
48
Abstract
This paper describes a methodology of creating custom dataset and then using convolution neural network (CNN) model for performing image processing operations on Google's t-rex game using tensorflow and keras framework. Machine learning professionals depend on online available datasets, specifically for computer vision based algorithms, since deep neural networks require specific input shape, size and channels before starting training process. Unfortunately, google's t-rex game does not have any directly available dataset unlike various image processing datasets. Therefore in this paper step by step process is provided for creating, training, and testing customized dataset with help of CNN model coded in python programming language for both GPU and CPU architectures.
Keywords
Computer Vision, Convolution Neural Networks (CNN), CPU, Customized Dataset, Deep Neural Networks, Google T-rex, GPU, Image Processing, Keras, Machine Learning, Python, Tensorflow
Department
Electrical Engineering
Recommended Citation
Dhananjai Bajpai and Lili He. "Custom Dataset Creation with Tensorflow Framework and Image Processing for Google T-Rex" 2020 12th International Conference on Computational Intelligence and Communication Networks (CICN) (2020): 45-48. https://doi.org/10.1109/CICN49253.2020.9242565
