Sketch recognition

About: Sketch recognition is a research topic. Over the lifetime, 1611 publications have been published within this topic receiving 40284 citations.

Flexible Parts-based Sketch Recognition

Abstract: Drawings and sketches are a natural way for people to communicate ideas, but it remains challenging to develop automated systems that can robustly recognize and interpret what is drawn. Most commonly, a drawing is first processed to obtain a low-level representation of that drawing in terms of lines or strokes, and this representation is then searched for matches to known object templates. In this chapter we propose two template-based methods for sketch recognition. A novel feature of these methods is that they both employ a hierarchy-of-parts template model that provides explicit support for templates with optional parts. This captures significant parts-based variation which would otherwise require a multitude of fixed-structure templates to model. The first method is developed for recognition in drawings consisting of sets of connected strokes and is applied as an interface for creating 3D models of airplanes, mugs, and fish. The second method allows for the recognition of more unstructured objects such as faces, plants, and sailboats in drawings that may also contain disjoint strokes. Neither method relies on the timing information of the input strokes, which may not be available for photographed or scanned drawings.

Skeletons from sketches of dancing poses

Abstract: The contribution of this paper is a sketch parser able to recognize the several components of a skeleton described using the drawing of a stick-man. We describe the sketch parser in detail, and briefly outline how it is applied to form the front-end of a sketch based retrieval system capable of searching for human poses in archival dance footage.

Sketch-Specific Data Augmentation for Freehand Sketch Recognition

Abstract: Sketch recognition remains a significant challenge due to the limited training data and the substantial intra-class variance of freehand sketches for the same object. Conventional methods for this task often rely on the availability of the temporal order of sketch strokes, additional cues acquired from different modalities and supervised augmentation of sketch datasets with real images, which also limit the applicability and feasibility of these methods in real scenarios. In this paper, we propose a novel sketch-specific data augmentation (SSDA) method that leverages the quantity and quality of the sketches automatically. From the aspect of quantity, we introduce a Bezier pivot based deformation (BPD) strategy to enrich the training data. Towards quality improvement, we present a mean stroke reconstruction (MSR) approach to generate a set of novel types of sketches with smaller intra-class variances. Both of these solutions are unrestricted from any multi-source data and temporal cues of sketches. Furthermore, we show that some recent deep convolutional neural network models that are trained on generic classes of real images can be better choices than most of the elaborate architectures that are designed explicitly for sketch recognition. As SSDA can be integrated with any convolutional neural networks, it has a distinct advantage over the existing methods. Our extensive experimental evaluations demonstrate that the proposed method achieves the state-of-the-art results (84.27%) on the TU-Berlin dataset, outperforming the human performance by a remarkable 11.17% increase. Finally, more experiments show the practical value of our approach for the task of sketch-based image retrieval.

Neural networks in off-line text recognition: a review

Abstract: This paper provides a brief review of the state-of-the-art of neural networks in off-line text recognition. We discuss the role that neural networks have played in text recognition. We also assess the state of the art of neural networks in character and word recognition. Despite the success of neural networks in character and word recognition, there are still many challenging problems.

Pattern Recognition Driven by Domain Ontologies

Abstract: The idea of using ontological models as a source of knowledge necessary to construct composite, multi-step pattern recognition procedures is presented in the paper. Special attention is paid to the structure of pattern recognition processes as sequences of decisions forming paths in bi-partite graphs describing pattern recognition networks. It is shown that such processes correspond to multi-step pattern recognition procedures described in literature, as well as they may describe more general classes of pattern recognition procedures. Construction of a multi-step pattern recognition procedure aimed at extraction of information about moving objects is illustrated by an example.