Geo-Spatiotemporal Features and Shape-Based Prior Knowledge for Fine-grained Imbalanced Data Classification

TL;DR

This paper introduces a novel approach for fine-grained wildlife classification by integrating geo-spatiotemporal features and shape-based prior knowledge to address challenges of small inter-class and large intra-class variations.

Contribution

It combines geo-spatiotemporal data and shape priors with existing methods to enhance fine-grained classification of imbalanced wildlife datasets.

Findings

Improved classification accuracy with geo-spatiotemporal features

Effective handling of imbalanced data in fine-grained classification

Enhanced recognition of subtle differences in wildlife images

Abstract

Fine-grained classification aims at distinguishing between items with similar global perception and patterns, but that differ by minute details. Our primary challenges come from both small inter-class variations and large intra-class variations. In this article, we propose to combine several innovations to improve fine-grained classification within the use-case of wildlife, which is of practical interest for experts. We utilize geo-spatiotemporal data to enrich the picture information and further improve the performance. We also investigate state-of-the-art methods for handling the imbalanced data issue.