Stroke-based sketched symbol reconstruction and segmentation
Hand-drawn objects usually consist of multiple semantically meaningful parts. For example, a stick figure consists of a head, a torso, and pairs of legs and arms. Efficient and accurate identification of these subparts promises to significantly improve algorithms for stylization, deformation, morphing and animation of 2D drawings. In this paper, we propose a neural network based model that segments symbols into stroke level components. Our segmentation network has two main elements: a fixed feature extractor and a multilayer perceptron (MLP) network that identifies a component based on the feature. As the feature extractor we utilize an encoder of the generative Variational Auto-Encoder (VAE) framework that draws symbols in stroke by stroke basis. Experiments show that a single encoder could be reused for segmenting multiple categories of sketches with negligible effects on segmentation accuracies. Extensive evaluations on our newly annotated dataset demonstrate that our model obtains significantly better scores as compared to the best baseline model. Moreover, our segmentation accuracies surpass existing methodologies on the available state of the art dataset. We also release our labeled dataset to the community.
READ FULL TEXT