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
TL;DR: Light-SRNet as discussed by the authors proposed a dual-attention mechanism to improve the accuracy of sketch recognition while retaining its lightweight nature, which achieved a recognition accuracy of 73.14% while requiring only about a quarter of the model parameters.
Abstract: Free-hand sketches play a vital role in graphically portraying ideas and concepts in image recognition systems. Most recently proposed learning-based sketch recognition methods have achieved marked progress in recognition accuracy, but they rarely optimize the use of the sparsity features of sketch images. Although several attention-based sketch recognition models have been presented, they endure complex computations and large model sizes. To address these challenges, we present a lightweight convolutional neural network called Light-SRNet based on a dual-attention mechanism to improve the accuracy of sketch recognition while retaining its lightweight nature. In the proposed model, we introduced both the spatial and channel attention mechanisms into the feature extraction network to highlight more discriminative feature representations to enhance its powerful sketch recognition ability. We compared the proposed Light-SRNet with its competitors on the TU-Berlin dataset, Sketchy dataset, and QuickDrawExtended dataset. Extensive experimental results show that Light-SRNet achieves a recognition accuracy of 73.14%, which is comparable to other similar sketch recognition techniques, while requiring only about a quarter of the model parameters.
01 Jan 2023
TL;DR: In this paper , a system that recognizes a few hand gestures and produces commands for human-computer interaction is presented, which is based on image processing and extraction techniques, followed by the project execution route followed by image processing.
Abstract: Gesture recognition systems reflect the user’s expressions in the real world, visually interpreting and incorporating them as a human–computer interaction channel. Recently, the demand for interaction by gesture has increased manifold and may ultimately replace the concept of mouse and keyboard, possibly soon. This has led to increased research in the area concerned with computer vision-based interpretation of hand gestures. The present work aims to develop a system that recognizes a few hand gestures and produces commands for human–computer interaction. The project execution route followed is image processing and extraction techniques.
26 Aug 2004
TL;DR: A method of optical formula recognition is described that consists of two major steps, namely, symbol recognition and structural analysis, which works reliably on almost noiseless images obtained by scanning among the usual documents clearly printed.
Abstract: Automatic recognition of formulas is one of the key parts in an OCR system. It could be really useful to be able to re-use knowledge in the scientific books which are not available in electronic form. A method of optical formula recognition is described. It consists of two major steps, namely, symbol recognition and structural analysis. Firstly, the search and process connect the components to gain the symbol components followed by symbol recognition. After that, we analyze the structure of the formula on the basis of the recognition result and the geometry features. The system works reliably on almost noiseless images obtained by scanning among the usual documents clearly printed.
Journal Article•
TL;DR: A hybrid SVM-HMM algorithm for sketch recognition is presented, which is designed to identify the sketches to the programming flowchart, and converts it to C language programs finally.
Abstract: The electronic white board and the tablet PC are bringing new technologies to modern education. This paper presents a hybrid SVM-HMM algorithm for sketch recognition. In this algorithm, ICA is used to reduce the dimension of features, a set of fuzzy SVMs are used as preliminary feature classifiers to produce fix length feature vector, which acts as a probability evaluator in the hidden states of Hidden Markov Models, and HMMs are employed as finally classifiers to recognize the unknown pattern. Experiment results show the hybrid algorithm has good learning and recognition ability. And based on this algorithm, a programming teaching system is designed to identify the sketches to the programming flowchart, and converts it to C language programs finally.
07 Jan 2012
TL;DR: The purpose of this evaluation is to establish the meadow of shrug recognition as a mechanism for interface with computers.
Abstract: Gestures recognition is one of the major areas of explore for the engineers, scientists and bioinformatics. GR is the innate approach of Human. A Hand Gesture Recognition System to identify factual instant shrug in churned up environments. It will identify a rift of hand gestures. How to make out the silhouette gesture for novel human-computer interface exclusive of controller required instantaneous hand gesture detection system by using shape milieu toning. The shape context is taken as a source portrayal for shape matching. It can be regarded as a universal depiction descriptor to symbolize the distribution of points in a set with scale and rotation invariance. User could intermingle with computer program by performing body gesture instead of physical contact. The image of hand gesture was captured. The hand gesture image was transformed into proper instruction according to the shape information respectively. The purpose of this evaluation is to establish the meadow of shrug recognition as a mechanism for interface with computers.