English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the zaia usage as premium feature

ZAIA

Presents the pdf analysis as premium feature

PDF Analysis

Insights

Presents the summarisation as premium feature

Summarisation

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the access of premium content as premium feature

Premium Content

Zendy Tools

Zendy Open

Predicting a landslide susceptibility map (LSM) is essential for riskrecognition and disaster prevention. Despite the successful application ofdata-driven approaches for LSM prediction, most methods generally apply asingle global model to predict the LSM for an entire target region. However, inlarge-scale areas with significant environmental change, various parts of theregion hold different landslide-inducing environments, and therefore, should bepredicted with respective models. This study first segmented target scenariosinto blocks for individual analysis. Then, the critical problem is that in eachblock with limited samples, conducting training and testing a model isimpossible for a satisfactory LSM prediction, especially in dangerousmountainous areas where landslide surveying is expensive. To solve the problem,we trained an intermediate representation by the meta-learning paradigm, whichis superior for capturing information valuable for few-shot adaption from LSMtasks. We hypothesized that there are more general and vital conceptsconcerning landslide causes and are sensitive to variations in input features.Thus, we can quickly few-shot adapt the models from the intermediaterepresentation for different blocks or even unseen tasks using very fewexemplar samples. Experimental results on the two study areas demonstrated thevalidity of our block-wise analysis in large scenarios and revealed the topfew-shot adaption performances of the proposed methods.

Meta-learning an Intermediate Representation for Few-shot Block-wise Prediction of Landslide Susceptibility