Bio

Nathan Dadap is a PhD student in Professor Alexandra Konings’ Group in the Earth System Science Department at Stanford University. He is interested in using remote sensing to better understand peatland hydrology - an important control on fire risk and carbon emissions. Currently, Nathan is working on a research project relating soil moisture and fire in Equatorial Asia. Prior to graduate school, Nathan worked at the U.S. Environmental Protection Agency on hazardous waste issues. Nathan holds a BS in Applied Physics from Columbia University.

Contact

  • Academic
    ndadap@stanford.edu
    University - Student Department: Department of Earth System Science Position: Graduate

Additional Info

All Publications

  • Drainage Canals in Southeast Asian Peatlands Increase Carbon Emissions AGU Advances Dadap, N. C., Hoyt, A. M., Cobb, A. R., Oner, D., Kozinski, M., Fua, P. V., Rao, K., Harvey, C. F., Konings, A. G. 2021; 2 (1): 1-14

    View details for DOI 10.1029/2020AV000321

  • Promoting Connectivity of Network-Like Structures by Enforcing Region Separation. IEEE transactions on pattern analysis and machine intelligence Oner, D., Kozinski, M., Citraro, L., Dadap, N. C., Konings, A. G., Fua, P. 2021; PP

    Abstract

    We propose a novel, connectivity-oriented loss function for training deep convolutional networks to reconstruct network-like structures, like roads and irrigation canals, from aerial images. The main idea behind our loss is to express the connectivity of roads, or canals, in terms of disconnections that they create between background regions of the image. In simple terms, a gap in the predicted road causes two background regions, that lie on the opposite sides of a ground truth road, to touch in prediction. Our loss function is designed to prevent such unwanted connections between background regions, and therefore close the gaps in predicted roads. It also prevents predicting false positive roads and canals by penalizing unwarranted disconnections of background regions. In order to capture even short, dead-ending road segments, we evaluate the loss in small image crops. We show, in experiments on two standard road benchmarks and a new data set of irrigation canals, that convnets trained with our loss function recover road connectivity so well that it suffices to skeletonize their output to produce state of the art maps. A distinct advantage of our approach is that the loss can be plugged in to any existing training setup without further modifications.

    View details for DOI 10.1109/TPAMI.2021.3074366

    View details for PubMedID 33881988

  • Satellite soil moisture observatins predict burned area in Southeast Asian peatlands ENVIRONMENTAL RESEARCH LETTERS Dadap, N. C., Cobb, A. R., Hoyt, A. M., Harvey, C. F., Konings, A. G. 2019; 14

    View details for DOI 10.1088/1748-9326/ab3891

https://orcid.org/0000-0001-7586-6210