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 42
- Downloads 0
A Review on Machine Learning and Its Applications
Learning is the most important aspect of human intellect and the most fundamental method of acquiring information. Machine learning is the most basic method for making a machine intelligent. Application Data has grown in...
Movie Recommendation System Using Item Based Collaborative Filtering
In today's digital world where there is an endless variety of content consumed such as books, videos, articles, Films, etc., finding material of one's choice has become an infallible task. Digital content on the other ha...
Behavior of Plastic Working in Oil-Field Equipment
Researches were made for the purpose of increase of labor productivity and replace of nonferrous metals and ferrous metals by plastics. Details for this purpose are selected with a certain characteristic that further the...
Syllabic Units Automatically Segmented Data for Continuous Speech Recognition
We present novel approach for constant speech processing in which the detection and recognition tasks are separated A syllable is utilized as a measure both to detection and localization. A minimal phase’s group delay ch...
CNC Machine Technologies: A Review
Portable, interoperable, and flexible are the objectives of following generations of computer-controlled technologies G-codes have long been used by CNC production instruments for component programmers and are now seen a...