Confidence Estimation for Object Detection in Document Images

TL;DR

This paper introduces four confidence estimators for object detection in document images, enhancing active learning by selecting more informative data and reducing computational costs.

Contribution

It proposes novel confidence estimators, including a cost-effective one based on descriptive statistics, improving active learning for document image analysis.

Findings

Three estimators significantly improve detection performance over random selection.

The descriptive statistics estimator can replace MC dropout without performance loss.

The methods reduce annotation effort and computational costs.

Abstract

Deep neural networks are becoming increasingly powerful and large and always require more labelled data to be trained. However, since annotating data is time-consuming, it is now necessary to develop systems that show good performance while learning on a limited amount of data. These data must be correctly chosen to obtain models that are still efficient. For this, the systems must be able to determine which data should be annotated to achieve the best results. In this paper, we propose four estimators to estimate the confidence of object detection predictions. The first two are based on Monte Carlo dropout, the third one on descriptive statistics and the last one on the detector posterior probabilities. In the active learning framework, the three first estimators show a significant improvement in performance for the detection of document physical pages and text lines compared to a random selection of images. We also show that the proposed estimator based on descriptive statistics can replace MC dropout, reducing the computational cost without compromising the performances.