Learning to segment on tiny datasets: a new shape model
This addresses the problem of limited data availability for segmentation tasks, offering a potential solution for scenarios where large datasets are impractical, though it appears incremental as it builds on existing part-based and CRF approaches.
The paper tackles object segmentation with tiny datasets by proposing a part-based algorithm using a novel shape descriptor and dense CRF model, achieving results near state-of-the-art big data methods in some cases.
Current object segmentation algorithms are based on the hypothesis that one has access to a very large amount of data. In this paper, we aim to segment objects using only tiny datasets. To this extent, we propose a new automatic part-based object segmentation algorithm for non-deformable and semi-deformable objects in natural backgrounds. We have developed a novel shape descriptor which models the local boundaries of an object's part. This shape descriptor is used in a bag-of-words approach for object detection. Once the detection process is performed, we use the background and foreground likelihood given by our trained shape model, and the information from the image content, to define a dense CRF model. We use a mean field approximation to solve it and thus segment the object of interest. Performance evaluated on different datasets shows that our approach can sometimes achieve results near state-of-the-art techniques based on big data while requiring only a tiny training set.