The paper “TrueBranch: Metric Learning-based Verification of Forest Conservation Projects” wins the Best Proposal Paper Award at the ICLR Workshop on Tackling Climate Change with ML

The paper “TrueBranch: Metric Learning-based Verification of Forest Conservation Projects” by Simona Santamaria (ETH Zurich); David Dao (ETH Zurich); Björn Lütjens (MIT); Ce Zhang (ETH Zurich) has won the Best Proposal Paper Award at the ICLR Workshop on Tackling Climate Change with ML.

The paper addresses the huge problem of deforestation that accounts for 15% of all global greenhouse gas emissions! This problem is not new and UN and companies have started mitigation programs that pay landowners if they can prove that they're conserving forests. However, proving the conservation of forests is currently done by on-ground inspections, which is so expensive that many non-profits and indigenous groups are excluded from these projects. So, in our paper, we propose drone-based and AI-verified inventories of forests to create a cheap and trustworthy monitoring system. If deployed this system could incentivize many more landowners to conserve forests.