scispace - formally typeset
Search or ask a question
Topic

Sketch recognition

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


Papers
More filters
Journal ArticleDOI
TL;DR: The evaluations have shown that Mechanix is as effective as paper-and-pencil-based homework for teaching method of joints truss analysis; focus groups with students who used the program have revealed that they believe Mechanix enhances their learning and that they are highly engaged while using it.
Abstract: Massive open online courses, online tutoring systems, and other computer homework systems are rapidly changing engineering education by providing increased student feedback and capitalizing upon online systems' scalability. While online homework systems provide great benefits, a growing concern among engineering educators is that students are losing both the critical art of sketching and the ability to take a real system and reduce it to an accurate but simplified free-body diagram (FBD). For example, some online systems allow the drag and drop of forces onto FBDs, but they do not allow the user to sketch the FBDs, which is a vital part of the learning process. In this paper, we discuss Mechanix, a sketch recognition tool that provides an efficient means for engineering students to learn how to draw truss FBDs and solve truss problems. The system allows students to sketch FBDs into a tablet computer or by using a mouse and a standard computer monitor. Using artificial intelligence, Mechanix can determine not only the component shapes and features of the diagram but also the relationships between those shapes and features. Because Mechanix is domain specific, it can use those relationships to determine not only whether a student's work is correct but also why it is incorrect. Mechanix is then able to provide immediate, constructive feedback to students without providing final answers. Within this manuscript, we document the inner workings of Mechanix, including the artificial intelligence behind the scenes, and present studies of the effects on student learning. The evaluations have shown that Mechanix is as effective as paper-and-pencil-based homework for teaching method of joints truss analysis; focus groups with students who used the program have revealed that they believe Mechanix enhances their learning and that they are highly engaged while using it.

23 citations

Proceedings ArticleDOI
24 Aug 2014
TL;DR: This paper proposes an online gesture recognition method for multimodal RGB-D data that achieves an 85% recognition accuracy with 20 gesture classes and can perform the recognition in real-time.
Abstract: Gesture recognition using RGB-D sensors has currently an important role in many fields such as human-computer interfaces, robotics control, and sign language recognition. However, the recognition of hand gestures under natural conditions with low spatial resolution and strong motion blur still remains an open research question. In this paper we propose an online gesture recognition method for multimodal RGB-D data. We extract multiple hand features with the assistance of body and hand masks from RGB and depth frames, and full-body features from the skeleton data. These features are classified by multiple Extreme Learning Machines on the frame level. The classifier outputs are then modeled on the sequence level and fused together to provide the final classification results for the gestures. We apply our method on the ChaLearn 2013 gesture dataset consisting of natural signs with the hand diameters in the images around 20-10 pixels. Our method achieves an 85% recognition accuracy with 20 gesture classes and can perform the recognition in real-time.

23 citations

Journal ArticleDOI
TL;DR: The extensive evaluation on diagrams from six different domains has shown that the resulting dividers, using LADTree and LogitBoost, are significantly more accurate than three existing dividers.

23 citations

Proceedings Article
09 Apr 2009
TL;DR: Hashigo, a kanji sketch interactive system which achieves human instructor-level critique and feedback on both the visual structure and written technique of students’ sketched kanji, is described.
Abstract: Language students can increase their effectiveness in learning written Japanese by mastering the visual structure and written technique of Japanese kanji. Yet, existing kanji handwriting recognition systems do not assess the written technique sufficiently enough to discourage students from developing bad learning habits. In this paper, we describe our work on Hashigo, a kanji sketch interactive system which achieves human instructor-level critique and feedback on both the visual structure and written technique of students’ sketched kanji. This type of automated critique and feedback allows students to target and correct specific deficiencies in their sketches that, if left untreated, are detrimental to effective long-term kanji learning.

23 citations

Proceedings ArticleDOI
04 Sep 2006
TL;DR: A parsing strategy for the recognition of hand-drawn diagrams that can be used in interactive sketch interfaces based on grammar formalism, namely sketch grammars (SkGs), for describing both the symbols' shape and the syntax of diagrammatic notations, and from which recognizers are automatically generated.
Abstract: The existing sketch recognizers perform only a limited drawing recognition since they process simple sketches, or rely on drawing style assumptions that reduce the recognition complexity, and in most cases they require a substantial amount of training data. In this paper we present a parsing strategy for the recognition of hand-drawn diagrams that can be used in interactive sketch interfaces. The approach is based on a grammar formalism, namely Sketch Grammars (SkGs), for describing both the symbols’ shape and the syntax of diagrammatic notations, and from which recognizers are automatically generated. The recognition system was evaluated in the domain of UML use case diagrams and the results highlight the recognition accuracy improvements produced by the use of context in the disambiguation process.

23 citations


Network Information
Related Topics (5)
Feature (computer vision)
128.2K papers, 1.7M citations
84% related
Object detection
46.1K papers, 1.3M citations
83% related
Feature extraction
111.8K papers, 2.1M citations
82% related
Image segmentation
79.6K papers, 1.8M citations
81% related
Convolutional neural network
74.7K papers, 2M citations
80% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
202326
202271
202130
202029
201946
201827