Research map / Detection and segmentation / Open-set and few-shot detection
GeoDE: a Geographically Diverse Evaluation Dataset for Object Recognition
Vikram V. Ramaswamy, Sing Yu Lin, Dora Zhao, Aaron Bryan Adcock, Laurens van der Maaten, Deepti Ghadiyaram, Olga Russakovsky
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add GeoDE images to ImageNet training data (same total size) | linear model on SwAV ResNet50 features · GeoDE test overall | accuracy (%): 67 -> 88 | Improved |
| add GeoDE images to ImageNet training data | linear model on SwAV ResNet50 features · DollarStreet overall | accuracy (%): 60 -> 69 | Improved |
| add GeoDE images to training | linear model · DollarStreet Africa | accuracy (%): 45 -> 55 | Improved |
| add GeoDE images to training | linear model · DollarStreet Europe | accuracy (%): 75 -> 80 | Improved |
| add GeoDE images to training | linear model · DollarStreet clean. equip. | accuracy (%): 19 -> 36 | Improved |
| add GeoDE images to training | linear model · DollarStreet dustbin | accuracy (%): 76 -> 68 | Hurt |
| add GeoDE images to training | linear model · DollarStreet stove | accuracy (%): 89 -> 85 | Hurt |
| add GeoDE images to training | linear model · DollarStreet light fixture | accuracy (%): 36 -> 63 | Improved |
| add data from one GeoDE region back into ImageNet training | linear model · GeoDE regional and Europe test | accuracy | Improved |
| evaluate on Africa instead of Europe images | CLIP and ImageNet models · GeoDE | accuracy (%): 85.8 -> 78.7 | Hurt |
| use CLIP instead of ImageNet-trained model | zero-shot recognizer · GeoDE Africa | accuracy (%): 62.7 -> 78.7 | Improved |
Closest papers by meaning
- Does Progress On Object Recognition Benchmarks Improve Generalization on Crowdsourced, Global Data? (ICLR 2024 · similarity 0.89)
- OpenStreetView-5M: The Many Roads to Global Visual Geolocation (CVPR 2024 · similarity 0.86)
- SELECT: A Large-Scale Benchmark of Data Curation Strategies for Image Classification (NeurIPS 2024 · similarity 0.85)
- COUNTS: Benchmarking Object Detectors and Multimodal Large Language Models under Distribution Shifts (CVPR 2025 · similarity 0.85)
- GeoRanker: Distance-Aware Ranking for Worldwide Image Geolocalization (NeurIPS 2025 · similarity 0.85)
- Benchmarking Object Detectors under Real-World Distribution Shifts in Satellite Imagery (CVPR 2025 · similarity 0.85)
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.
