EcoMatcher: Efficient Clustering Oriented Matcher for Detector-Free Image Matching

Detector-free local feature matching methods have demonstrated significant performance improvements since leveraging the power of Transformer architecture. The global receptive field allows for simultaneous interaction among all elements, proving particularly beneficial in regions with low texture or repetitive patterns. However, Transformer-based methods are confronted by how to achieve a balance between computational cost and expressive efficacy when dealing with numerous features. In this work, we revisit existing detector-free methods and propose EcoMatcher, a universal matcher based on implicit clustering termed Context Clusters. By introducing coarse-grained features as clustering centers, similar features are allocated to the same center, forming distinct clustering patterns. Features within the same cluster are then dispatched with identical messages from their center but at varying scales depending on the similarity metrics. This process defines a novel feature extraction paradigm for both self-understanding and cross-interaction of image pairs, facilitating fusing multi-level features and reducing overall complexity. EcoMatcher proves to be a competitive detector-free method in terms of memory consumption and runtime speed, while also achieves strong performance on mainstream benchmarks.