Deep Learning

Fine-Grained Species Recognition With Privileged Pooling: Better Sample Efficiency Through Supervised Attention

We propose a scheme that uses privileged information, in the form of keypoint annotations for the training data, to learn strong models from small and/or biased training sets.

Robust Damage Estimation of Typhoon Goni on Coconut Crops with Sentinel-2 Imagery

We evaluate the damages on coconut crops caused by Typhoon goni using Senitnel-2 imagery. Overall we estimated that 14.1 M coconut trees were affected by the typhoon inside our area of study.

Mapping Oil Palm Density at Country Scale: an Active Learning Approach

We propose a new, active deep learning method to estimate oil palm density at scale of from Sentinel-2 satellite images, and apply it to generate complete maps for Malaysia and Indonesia.

Recognition of Unseen Bird Species by Learning from Field Guides

We exploit existing field guides to learn bird species recognition in large scale datasets for zero-shot recognition of unseen species

Counting the uncountable: deep semantic density estimation from space

We propose a new method to count objects of specific categories, including oil palm trees, olive trees and cars. All of these objects are significantly smaller than the ground sampling distance of a satellite image.

Fast cosmic web simulations with generative adversarial networks

We demonstrate the application of a machine learning technique called Generative Adversarial Networks (GAN) to learn models that can efficiently generate new, physically realistic realizations of the cosmic web

Example Project

An example of using the in-built project page.