Sketch recognition

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

Research on Recognition of On-line Freehand Sketches

Abstract: As a direct,convenient,natural and efficient human-machine interactive way,freehand sketching becomes more and more important during the development of CAD technologiesAn intelligent freehand sketching system is described in order to realize to combine the flexibility and ease of paper and pencil with the processing power of computersThe system integrates several process like sketch preprocessing,feature points detection,strokes fragment,primitives recognition,shape reconstruction,and primitives beautificationThe result of experiments shows that the method has good recognition performance

A hierarchical residual network with compact triplet-center loss for sketch recognition

Abstract: With the widespread use of touch-screen devices, it is more and more convenient for people to draw sketches on screen. This results in the demand for automatically understanding the sketches. Thus, the sketch recognition task becomes more significant than before. To accomplish this task, it is necessary to solve the critical issue of improving the distinction of the sketch features. To this end, we have made efforts in three aspects. First, a novel multi-scale residual block is designed. Compared with the conventional basic residual block, it can better perceive multi-scale information and reduce the number of parameters during training. Second, a hierarchical residual structure is built by stacking multi-scale residual blocks in a specific way. In contrast with the single-level residual structure, the learned features from this structure are more sufficient. Last but not least, the compact triplet-center loss is proposed specifically for the sketch recognition task. It can solve the problem that the triplet-center loss does not fully consider too large intra-class space and too small inter-class space in sketch field. By studying the above modules, a hierarchical residual network as a whole is proposed for sketch recognition and evaluated on Tu-Berlin benchmark thoroughly. The experimental results show that the proposed network outperforms most of baseline methods and it is excellent among non-sequential models at present.

Multi-domain Hierarchical Free-Sketch Recognition Using Graphical Models

Abstract: In recent years there has been an increasing interest in sketch-based user interfaces, but the problem of robust free-sketch recognition remains largely unsolved. This chapter presents a graphical-model-based approach to free-sketch recognition that uses context to improve recognition accuracy without placing unnatural constraints on the way the user draws. Our approach uses context to guide the search for possible interpretations and uses a novel form of dynamically constructed Bayesian networks to evaluate these interpretations. An evaluation of this approach on two domains—family trees and circuit diagrams—reveals that in both domains the use of context to reclassify low-level shapes significantly reduces recognition error over a baseline system that does not reinterpret low-level classifications. Finally, we discuss an emerging technique to solve a major remaining challenge for multi-domain sketch recognition revealed by our evaluation: the problem of grouping strokes into individual symbols reliably and efficiently, without placing unnatural constraints on the user’s drawing style.

Enhanced Autonomous Face Recognition Machine. Volume 1

Abstract: : This thesis continues work on the Autonomous Face Recognition Machine developed at AFIT in 1985. There were two major changes made to the system. The set of features extracted from the face for use in the recognition process, was changed. A higher dimensioned vector taken from the two-dimensional Discrete Fourier Transform of the face, was used in hope of increasing the separation of templates stored in the data base. Further research is needed to determine whether this change is beneficial to the system. The second change was to the decision rule used in recognition. The decision making portion of the system was replaced by a back propagation neural network. While providing equivalent recognition capability, this change provides a constant recognition time independent of the number of subjects trained into the system. Keywords: Pattern recognition, Image processing, Neural networks, Gesalt transforms.