AmalgaMatch: 187 Multimodal Microscopy Image Pairs for Materials Science

187 image pairs from the AmalgaMatch dataset, partitioned into six distinct matching tasks and 19 material subsets, facilitate evaluation of foundation models for multimodal image registration. Ali Riza Durmaz published this supplementary PDF in May 2026 under a CC-BY-4.0 license. The dataset covers metals, alloys, and ceramics imaged with diverse microscopy modalities, presenting challenges like limited mutual information and field-of-view ratios as low as 2%.

Use Cases

• Benchmarking foundation models like MatchAnything RoMa and ELoFTR on out-of-domain scientific imagery based on the described 187 image pairs.
• Developing inference strategies for image matching tasks with strong field-of-view mismatch, as mentioned for dislocation characterization groups.
• Training or fine-tuning multimodal matchers for correlative microscopy workflows based on the six distinct task groups described.
• Studying model generalization on image pairs with limited mutual information and distinct length scales, a noted challenge in materials science.

Strengths

• Contains 187 image pairs organized into six distinct matching task groups, providing structured evaluation scenarios.
• Covers 19 subsets representing different materials, including metals, alloys, and ceramics, offering diversity.
• Includes image pairs with pronounced field-of-view disparities, with ratios as low as 2%, representing real-world challenges.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The primary file is a 4.7 MB PDF, which likely contains analysis and figures rather than the raw image data itself.

Provenance

Source

figshare, authored by Ali Riza Durmaz.

Collection Method

Likely compiled from correlative materials microscopy experiments, as referenced in the description.

Time Range

null

Freshness

Last updated 2026-05-21 05:43:58; freshness should be verified.

Geography

null