Search papers, labs, and topics across Lattice.
This paper introduces RouteGraph-Mona, an enhanced version of the Multi-cognitive Visual Adapter (Mona) that addresses the limitations of static multi-scale aggregation in mineral image classification. By implementing sample-adaptive routing and regularizing routing signatures with class-wise anchors and confusion-weighted margins, RouteGraph-Mona effectively captures scale preferences and improves class separation among visually similar minerals. Experimental results demonstrate that RouteGraph-Mona outperforms the original Mona model in mean accuracy across three public datasets while remaining competitive with established fine-tuning methods.
RouteGraph-Mona achieves superior mineral classification accuracy by dynamically adapting routing based on sample-specific scale preferences, addressing a critical challenge in geological imaging.
Mineral image classification is important for geological exploration and resource development, but it remains challenging due to substantial intra-class variations in appearance and high inter-class visual similarity. Multi-cognitive Visual Adapter (Mona) is a vision-oriented parameter-efficient adapter that adapts pre-trained visual models by tuning only a few parameters. However, Mona statically aggregates responses from multiple scales, limiting its ability to accommodate sample-specific scale preferences and model confusion among visually similar mineral categories. To address this issue, we propose \textbf{RouteGraph-Mona}, a lightweight route-space regularization method built on Mona. Specifically, we replace Mona's static multi-scale aggregation with sample-adaptive routing. The resulting branch-selection behavior defines a compact routing space that captures each image's scale preferences. We then regularize the resulting routing signatures with class-wise route anchors and confusion-weighted margins. The route anchors encourage class-consistent routing patterns, while the margins promote greater separation between visually similar categories in the routing space. Experiments on three public mineral image datasets with two visual backbones show that RouteGraph-Mona consistently outperforms Mona in mean accuracy and remains competitive with representative fine-tuning methods and mineral image classification baselines.
