AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

SatBird: a Dataset for Bird Species Distribution Modeling using Remote Sensing and Citizen Science Data

Mélisande Teng, Amna Elmustafa, Benjamin Akera, Yoshua Bengio, Hager Radi, Hugo Larochelle, David Rolnick

NeurIPS 2023 · Probabilistic and scientific modeling › Molecules and proteins · approach: Statistical analysis of ML practice · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add environmental rasters to RGB image input (ablation)ResNet-18 · SatBird-USA-summerTop-k accuracy: 54.5 -> 66.9Improved
add environmental rasters to RGB image input (ablation)ResNet-18 · SatBird-USA-winterTop-k accuracy: 56.9 -> 70.9Improved
use RGBNIR reflectance instead of RGB image (no env data) (ablation)ResNet-18 · SatBird-USA-summerTop-10 accuracy: 35.5 -> 36.1Improved
image-only ResNet-18 compared with environmental baselineResNet-18 (img) · SatBird-USA-summerTop-k accuracy: 63.8 -> 54.5Hurt
add range maps (ablation)ResNet-18 (refl+env) · SatBird-USA-summerTop-k accuracy: 66.6 -> 66.7No clear effect
add range maps (ablation)ResNet-18 (refl+env) · SatBird-US-summer subsetTop-k accuracy: 72.9 -> 73.4Improved
add location encoder instead of range maps (ablation)ResNet-18 (refl+env) · SatBird-US-summer subsetTop-k accuracy: 72.9 -> 72.0Hurt
replace ImageNet initialization with Seasonal Contrast weights (ablation)ResNet-18 (refl+env+RM) · SatBird-USA-summerTop-k accuracy: 66.8 -> 66.6No clear effect
replace cross entropy loss with L1 loss (ablation)ResNet-18 (refl+env+RM) · SatBird-USA-summerTop-k accuracy: 66.78 -> 61.13Hurt
replace cross entropy loss with focal loss (ablation)ResNet-18 (refl+env+RM) · SatBird-USA-summerTop-k accuracy: 66.78 -> 60.68Hurt
replace cross entropy loss with L1 loss (ablation)ResNet-18 (refl+env+RM) · SatBird-USA-summerMAE [1e-2]: 2.16 -> 1.86Improved
use frozen pretrained Satlas instead of ResNet-18 on imageSatlas · SatBird-USA-summerTop-10 accuracy: 35.5 -> 30.5Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy