Summary of Kidsat: Satellite Imagery to Map Childhood Poverty Dataset and Benchmark, by Makkunda Sharma et al.
KidSat: satellite imagery to map childhood poverty dataset and benchmark
by Makkunda Sharma, Fan Yang, Duy-Nhat Vo, Esra Suel, Swapnil Mishra, Samir Bhatt, Oliver Fiala, William Rudgard, Seth Flaxman
First submitted to arxiv on: 8 Jul 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed paper addresses the challenge of developing a standardized benchmark for evaluating deep learning models on demographic, health, and development indicators using satellite imagery. The authors introduce a new dataset pairing satellite imagery with high-quality survey data on child poverty, which is calculated from UNICEF’s multidimensional child poverty framework. The dataset consists of 33,608 images from 19 countries in Eastern and Southern Africa, spanning the period from 1997 to 2022. The paper tests spatial and temporal generalization by evaluating models on unseen locations and data after the training years. Multiple models are benchmarked, including low-level satellite imagery models like MOSAIKS and deep learning foundation models such as DINOv2 and SatMAE. The authors provide open-source code for building the satellite dataset, obtaining ground truth data from DHS, and running various models assessed in their work. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about creating a new way to measure child poverty using pictures taken from space. It’s like a puzzle where you need to put together different pieces of information to get an accurate picture. The researchers made a big dataset with 33,608 images that show how poor or not poor children are in 19 countries. They tested different computer models to see which ones work best for this task. They want to help people understand better and make decisions based on the data they collect from space. |
Keywords
» Artificial intelligence » Deep learning » Generalization
