Welcome to the cutting-edge materials science! At the heart of our project lies the **Materials Data Segmentation Benchmark** and the **Materials Data Segmentation Model Garden**, two pioneering initiatives designed to revolutionize the evaluation and development of materials segmentation models. ***All content on this page is open-source and everyone can contribute.*** ***Want to contribute? Fill up the [Google Form][1].*** <h2> Materials Data Segmentation Benchmark </h2> <p> MDSB is our central repository, offering an extensive collection of diverse materials datasets. From metals and alloys to polymers, our datasets are available in both 2D and 3D formats, complete with ground truth annotations. We support a variety of imaging modalities, including microscopy and tomography. As an open-source initiative, MDSB is a collaborative platform that fosters innovation and advancement in materials analysis techniques. </p> <h2> Materials Data Segmentation Model Garden </h2> <p> Complementing MDSB, MDSMG provides a standardized platform for testing and comparing segmentation models. Our focus on evaluation metrics ensures consistent performance assessment across models. As an open-source project, MDSMG democratizes access to powerful segmentation tools, accelerating the development of new materials analysis techniques. </p> <h3> Our Vision </h3> Together, MDSB and MDSMG form a comprehensive ecosystem that supports the evolution of materials segmentation research. We aim to provide a robust benchmarking infrastructure and a collaborative space for model testing and development in scientific and engineering domains. <h3> Our Datasets and Segmentation Models </h3> We have meticulously curated the initial collection of images from various datasets, including AFM, SEM, Radiography, and XCT. Each dataset comprises a mix of hardness levels, ranging from very easy to very hard, with 20% of images in each category. This diverse collection enables comprehensive evaluation and comparison of the performance of the segmentation models on the materials science specific tasks. <br /> <br /> Currently, we have organized six different datasets across various characterization techniques and materials. These datasets are stored in separate components. 1. [LPBF (High-Speed Camera)][2] 2. [Fractography (SEM)][3] 3. [Crystallites (AFM)][4] 4. [Al-Mg-Si Stress corrosion (XCT)][5] 5. [Al Pitting corrosion (XCT)][6] 6. [Contact corrosion (SEM)][7] <br /> Furthermore, for each dataset, we provide raw images, corresponding ground truth images, models, and evaluation scripts, all organized in separate folders. These evaluation scripts can be used to assess the models on their respective benchmark datasets. <br /> Our plan is to vastly expand our collection, aiming to expand this to include new image types, new models that cover both 2D and 3D data. This will provide an unparalleled resource for comprehensive training and development of new segmentation models, pushing the boundaries of materials science. <h3> Join Us </h3> By participating in our open-source initiatives, you become part of a global community dedicated to pushing the boundaries of materials analysis. Collaborate with us, contribute to our repositories, and help shape the future of materials segmentation research. <br> <br /> You can fill out this [Google Form][8] to contribute your datasets and models. Let's work together to advance the frontiers of materials science. <br /> Explore, innovate, and lead the way with MDSB and MDSMG. Together, we can unlock new possibilities in materials science and engineering. ---------- For more details, and collaboration opportunities, visit [our website][9]. [1]: https://forms.gle/Jduxw7wQfrsxCjhA7 [2]: https://osf.io/dzv3b/ [3]: https://osf.io/gdwyb/ [4]: https://osf.io/b2wdg/ [5]: https://osf.io/kt9wx/ [6]: https://osf.io/uzxpc/ [7]: https://osf.io/mwg6n/ [8]: https://forms.gle/6fW9To6EhdupWvks7 [9]: https://mds3-coe.com/research/materials-data-segmentation-benchmark-mdsb-and-model-garden-msdmg/