Resizing Tiny Imagenet: An Iterative Approach Towards Image Classification
Journal Title: International Journal of Innovative Research in Computer Science and Technology - Year 2020, Vol 8, Issue 6
Abstract
Deep neural networks have attained almost human-level performance over several Image classification and object detection problems, deploying robust and powerful state-of-the-art networks. Stanford’s Tiny ImageNet dataset has been around for a while and neural networks have struggled to classify them. It is generally considered one of the harder datasets in the domain of image classification. The validation accuracy of the existing systems maxes out at 61- 62% with a select few shooting beyond 68-69%. These approaches are often plagued by problems of overfitting and vanishing gradients. In this paper, we present a new method to get above average validation accuracy while circumventing these problems. We use the resizing image technique which trains multiple model times over different image sizes. This approach enables the model to derive more relevant and precise features as we move towards the original image size. It also makes the training process adaptable to available hardware resources. After reaching the image size, we hyper-tune other parameters such as Learning Rate, Optimizers, etc. to increase the overall validation accuracy. The final validation accuracy of the model using resizing and hyper tuning is 62.57%.
Authors and Affiliations
Praveen Kumar, Rishabh Singh, Nilesh Kumar Singh, Hitesh Agarwal
Keywords
Related Articles
- EP ID EP747612
- DOI 10.21276/ijircst.2020.8.6.8
- Views 23
- Downloads 0
The Variational Methods for Solving Random Models
This paper studies the solutions of variational methods for random ordinary (partial) dierential equations in L2−space. These methods are called Galerkin method, Petrov-Galerkin method, Least-Squares method and Collocati...
Microgrids Based on Regionalised Authority and Control
Since its inception in the late 1800s, the electrical infrastructure has only grown throughout the previous century. The demand for and need for this infrastructure will increase as the world's reliance on electrical ene...
A Review on Multilevel Approaches for Security in Cloud by Using ABE
Public Key Encryption acts as the basic technique for ABE where it provides one to many encryptions, here, the private key of users & the ciphertext both rely on attributes such that, when the set of the attributes of us...
Anomaly Detection in Credit Card Transactions using Machine Learning
Anomaly Detection is a method of identifying the suspicious occurrence of events and data items that could create problems for the concerned authorities. Data anomalies are usually associated with issues such as security...
Prediction of Weather Forecasting by Using Machine Learning
Weather forecasting has become important now a day because of varying climatic conditions around the world. Many technologies have been introduced to predict the weather whose accuracy is around 70%.Weather forecasting i...