cigCK: A data of 3D synthetic seismic volumes with labeled channels and karsts for deep-learning-based simultaneous channel and karst interpretation

cigCK is a dataset created by the Computational Interpretation Group (CIG) for the deep-learning-based simultaneous channel and karst interpretation in 3D seismic images. Jiarun Yang, Guangyu Wang, and Xinming Wu are the main contributors to the dataset. This dataset contains 200 pairs of synthetic 3D seismic images and the corresponding label images with the ground truth of channels and karsts simulated in the seismic images. More detail of building this dataset is discussed in the paper submitted to the journal of JGR Solid Earth: Yang, J., X. Wu* and G. Wang, 2025, Deep learning for simultaneous karst and channel detectionIEEE GRSL, accepted. Below are some brief descriptions of the dataset: The "gx" data is split into multiple parts due to file size limitations. Specifically, it contains gx.zip.001 – gx.zip.011. Please note that the file gx.zip.010 is hosted separately at the following Zenodo record: https://zenodo.org/records/17104253. All parts need to be downloaded and concatenated to reconstruct the full gx dataset, which includes 200 3D seismic volumes, each with dimensions of 500×500×250. The "lx.zip" file contains 200 corresponding label volumes, also with dimensions of 500×500×250. In a label image, areas representing karsts are assigned a value of 1, channels are denoted with a 2, and all other regions are marked as 0, which is why the compressed label images in lx.zip are much smaller than the seismic images compressed in gx.