M. Quamer Nasim is an AI/ML research engineer with the DK AI Holding group (since 2020), working on EarthScan's geoscience AI. He holds a PhD in Geophysics and Seismology from IIT Kharagpur. His work centres on multi-GPU deep learning for borehole images, well logs and seismic facies: fracture and bedding detection, vug quantification, raster well-log digitisation and well-to-well correlation.
He is first author of EarthAdaptNet (IEEE TGRS, 2022), VeerNet (Journal of Imaging, 2023), and the GeoBFDT and vug-quantification papers (SPE Journal, 2025). His published work covers:
- EarthAdaptNet (IEEE Transactions on Geoscience and Remote Sensing, 2022): seismic facies analysis that pairs domain adaptation with attention-based segmentation.
- VeerNet (Journal of Imaging, 2023): curve classification and digitisation of raster well-log images, the basis of the production ES Raster Digitizer pipeline.
- GeoBFDT (SPE Journal, 2025): automated detection of beddings and fractures in borehole images.
- Vug quantification (SPE Journal, 2025): adaptive thresholding, Gaussian weighting and Laplacian contrast analysis for vugs in borehole images.
- Automated well-log correlation using traditional + borehole image logs (EAGE Workshop on Borehole Geology in Asia Pacific): multi-modal log streams with deep image segmentation to detect formation boundaries, fractures and breakouts.