Scalable Model-Assisted Multi-Target Estimation in Large Image Collections
2026-07-20 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors study how to better combine computer vision model predictions with human labels to measure many things at once, like counting different object types in images. They build on past methods that gave good estimates for single numbers but extend the approach to handle multiple targets simultaneously. By testing different sampling techniques, they find some work better with limited human labeling than others depending on the situation. Their framework helps turn imperfect model guesses and a small number of human checks into reliable scientific measurements.
computer visionmulti-target estimationimportance samplinguniform samplingcontrol variatesratio estimatorMonte Carlo methodsannotation budgetsegmentationdetection datasets
Authors
Max Hamilton, Jinlin Lai, Daniel Sheldon, Subhransu Maji
Abstract
Computer vision models are increasingly used as measurement tools to estimate population-level quantities from large image collections, but prediction errors introduce bias and the resulting estimates lack statistical guarantees required in scientific applications. Prior work uses a Monte Carlo framework to combine model predictions with ground-truth annotations by sampling some images for humans to label and is able to provide unbiased estimates with controllable accuracy, but primarily addresses single-scalar estimation. We study the more general problem of multi-target estimation, where many quantities (e.g., class counts or proportions) must be estimated simultaneously, and adapt sampling and estimation strategies from survey sampling to this setting. Evaluations on five detection and segmentation datasets with 7-80 classes show that importance sampling excels with moderate annotation budgets or fewer targets, whereas uniform sampling with control variates is superior when estimating many targets or operating with minimal labels. Additionally, a subset-based ratio estimator remains highly competitive across all regimes. Ultimately, our framework effectively combines biased model predictions and limited human labels into rigorous scientific measurements.