Challenges in Fine-Grained Image Analysis

Serge Belongie

University of Copenhagen, DK

Abstract

Fine-grained image analysis (FGIA) is a longstanding and fundamental problem in computer vision and pattern recognition, and underpins a diverse set of real-world applications. The task of FGIA is concerned with visual objects from subordinate categories, e.g., species of birds or models of cars. The small inter-class and large intra-class variation inherent to fine-grained image analysis makes it a challenging problem. Capitalizing on advances in deep learning and large pre-trained multimodal models, in recent years we have witnessed remarkable progress in FGIA. In this talk we review representative examples in the context of recognition, retrieval, and generation/synthesis. In addition, we also review other key issues of FGIA, such as publicly available benchmark datasets, related domain-specific applications, and connections with other modalities including text and audio. We conclude by highlighting several research directions and open problems.