Cowbree: A novel dataset for fine-grained visual categorization

Umar Akbar Khan,Murtaja Ali Saare,Saima Anwar Lashari,Saira Moin U Din,Muhammad Ilyas

doi:10.11591/eei.v9i5.2443

Abstract

Fine-grained visual categorization (FGVC) dealt with objects belonging to one class with intra-class differences into subclasses. FGVC is challenging due to the fact that it is very difficult to collect enough training samples. This study presents a novel image dataset named Cowbreefor FGVC. Cowbree dataset contains 4000 images belongs to eight different cow breeds. Images are properly categorized under different breed names (labels) based on different texture and color features with the help of experts. While evidence shows that the existing dataset are of low quality, targeting few breeds with less number of images. To validate the dataset, three state of the art classifiers sequential minimal optimization (SMO), Multiclass classifier and J48 were used. Their results in term of accuracy are 68.81%, 55.81% and 57.45% respectively. Where results shows that SMO out performed with 68.81% accuracy, 68.4% precision and 68.8% recall.

Highlights

Classification task performed without deep knowledge of the domain is always challenging
Where results shows that sequential minimal optimization (SMO) out performed with 68.81% accuracy, 68.4% precision and 68.8% recall
In Computer vision Fine-grained visual categorization (FGVC) refers to categorize the objects belonging to one class but having the intra-class differences into subclasses

Summary

Introduction

Classification task performed without deep knowledge of the domain is always challenging It is more difficult when there are specific intra-class differences between objects. In Computer vision Fine-grained visual categorization (FGVC) refers to categorize the objects belonging to one class but having the intra-class differences into subclasses. Fine-grained visual categorization (FGVC) is challenging because, it is often difficult to acquire enough number of training samples [1,2]. The existing datasets in this scenario have few training samples in each category. Collection and classification of data for specific category of object is a very challenging task. It requires deep domain knowledge and experience. For this particular reason datasets for FGVC is very limited

Objectives

Results

Conclusion