Learning User's confidence for active learning
This work addresses a specific problem in remote sensing active learning by improving query efficiency for users, but it is incremental as it builds on existing active learning frameworks.
The paper tackles the conflict between active learning heuristics and user labeling confidence by proposing a filtering scheme that learns user confidence to minimize unanswerable queries, showing efficiency in maximizing useful queries compared to traditional methods on QuickBird images.
In this paper, we study the applicability of active learning in operative scenarios: more particularly, we consider the well-known contradiction between the active learning heuristics, which rank the pixels according to their uncertainty, and the user's confidence in labeling, which is related to both the homogeneity of the pixel context and user's knowledge of the scene. We propose a filtering scheme based on a classifier that learns the confidence of the user in labeling, thus minimizing the queries where the user would not be able to provide a class for the pixel. The capacity of a model to learn the user's confidence is studied in detail, also showing the effect of resolution is such a learning task. Experiments on two QuickBird images of different resolutions (with and without pansharpening) and considering committees of users prove the efficiency of the filtering scheme proposed, which maximizes the number of useful queries with respect to traditional active learning.