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 published on a yearly basis
Papers
More filters
••
14 Feb 2012TL;DR: This work discusses the design of PhysicsBook, a prototype system that enables users to solve physics problems using a sketch-based interface and then animates any diagram used in solving the problem to show that the solution is correct.
Abstract: We present PhysicsBook, a prototype system that enables users to solve physics problems using a sketch-based interface and then animates any diagram used in solving the problem to show that the solution is correct. PhysicsBook recognizes the diagrams in the solution and infers relationships among diagram components through the recognition of mathematics and annotations such as arrows and dotted lines. For animation, PhysicsBook uses a customized physics engine that provides entry points for hand-written mathematics and diagrams. We discuss the design of PhysicsBook, including details of algorithms for sketch recognition, inference of user intent and creation of animations based on the mathematics written by a user. Specifically, we describe how the physics engine uses domain knowledge to perform data transformations in instances where it cannot use a given equation directly. This enables PhysicsBook to deal with domains of problems that are not directly related to classical mechanics. We provide examples of scenarios of how PhysicsBook could be used as part of an intelligent tutoring system and discuss the strengths and weaknesses of our current prototype. Lastly, we present the findings of a preliminary usability study with five participants.
26 citations
••
03 Sep 2006TL;DR: This work draws on constellation models first proposed in the computer vision literature to develop probabilistic models for object sketches, based on multiple example drawings, which are applied to estimate the most-likely labels for a new sketch.
Abstract: Sketch-based modeling shares many of the difficulties of the branch of computer vision that deals with single image interpretation. Most obviously, they must both identify the parts observed in a given 2D drawing or image.We draw on constellation models first proposed in the computer vision literature to develop probabilistic models for object sketches, based on multiple example drawings. These models are then applied to estimate the most-likely labels for a new sketch. A multi-pass branch-and-bound algorithm allows well-formed sketches to be quickly labelled, while still supporting the recognition of more ambiguous sketches. Results are presented for five classes of objects.
26 citations
••
TL;DR: This review paper reconsiders the assumption of recognition as a pair-matching test, and introduces a new formal definition that captures the broader context of the problem, including how often metric properties are violated by recognition algorithms.
26 citations
••
22 Aug 2007TL;DR: A fast feature detection and description approach which can significantly speed up hand gesture recognition and obtains comparative performance with its counterpart in literatures.
Abstract: Hand gesture has been used as a natural and efficient way in human computer interaction. Due to independence of auxiliary input devices, vision-based hand interfaces is more favorable for users. However, the process of hand gesture recognition is very time consuming, which often brings much frustration to users. In this paper, we propose a fast feature detection and description approach which can significantly speed up hand gesture recognition. Firstly, integral image is used to approximate Gaussian derivatives to calculate image convolution in feature detection. Then multi-scale geometric descriptors at feature points are obtained to represent hand gestures. Finally gesture is recognized with its geometric configuration. Experiments show that the proposed method needs much less time consumption while obtains comparative performance with its counterpart in literatures.
26 citations
•
TL;DR: An overview of gesture recognition in real time using the concepts of correlation and Mahalanobis distance is presented and the six universal emotional categories namely joy, anger, fear, disgust, sadness and surprise are considered.
Abstract: Augmenting human computer interaction with automated analysis and synthesis of facial expressions is a goal towards which much research effort has been devoted recently. Facial gesture recognition is one of the important component of natural human-machine interfaces; it may also be used in behavioural science, security systems and in clinical practice. Although humans recognise facial expressions virtually without effort or delay, reliable expression recognition by machine is still a challenge. The face expression recognition problem is challenging because different individuals display the same expression differently. This paper presents an overview of gesture recognition in real time using the concepts of correlation and Mahalanobis distance.We consider the six universal emotional categories namely joy, anger, fear, disgust, sadness and surprise.
26 citations