- Research Article
- 10.70267/ic-aimees.202510
Existing Results and Recent Advancements of Fine-grained Image Classification
- Dec 19, 2025
- Exploring Science Academic Conference Series
- Weisheng Kong
Fine-grained image classification emphasizes the resolution of recognizing the subtle differences between subclass objects from the same category. Its classic examples of application include the classification of bird species, car brands and types of plants. During recent years, with the rapid popularization of machine learning and computer vision, there have been growing real circumstances concerning the issue of Fine-grained image classification, which greatly boosts the relevant research. However, there are few reviews of the existing results and recent advancements of Fine-grained image classification. This paper will systematically conclude the major research focuses of fine-grained image classification, highlighting two major aspects. The first aspect is fine-grained image classification based on feature fusion theory. The second aspect is fine-grained image classification based on an attention mechanism. Then, this paper introduces the collaborative application of feature fusion and attention mechanisms in fine-grained image classification. Next, the paper lists the mainstream data collection of fine-grained image classification. Lastly, this paper summarizes the main problems in current research and proposes future directions of this issue.
Read more