Abstract:Freehand sketching is a dynamic process where points are sequentially sampled and grouped as strokes for sketch acquisition on electronic devices. To recognize a sketched object, most existing methods discard such important temporal ordering and grouping information from human and simply rasterize sketches into binary images for classification. In this paper, we propose a novel single-branch attentive network architecture RNN-Rasterization-CNN (Sketch-R2CNN for short) to fully leverage the dynamics in sketches for recognition. Sketch-R2CNN takes as input only a vector sketch with grouped sequences of points, and uses an RNN for stroke attention estimation in the vector space and a CNN for 2D feature extraction in the pixel space respectively. To bridge the gap between these two spaces in neural networks, we propose a neural line rasterization module to convert the vector sketch along with the attention estimated by RNN into a bitmap image, which is subsequently consumed by CNN. The neural line rasterization module is designed in a differentiable way to yield a unified pipeline for end-to-end learning. We perform experiments on existing large-scale sketch recognition benchmarks and show that by exploiting the sketch dynamics with the attention mechanism, our method is more robust and achieves better performance than the state-of-the-art methods.

On-Line Sketch Recognition Using Direction Feature

Pen-based gesture design and recognition for sketch editing

An online sketching and gesture editing system for conceptual design

Sketch-R2CNN: an Attentive Network for Vector Sketch Recognition

An on-line sketch recognition algorithm for composite shape

<i>Sketch-R2CNN</i>: An RNN-Rasterization-CNN Architecture for Vector Sketch Recognition

Online Sketch Recognition Using Geometric-Based Classifiers

Recognition of On-line Sketched Electrical Diagrams

Classified Segmentation of Online Complex Freehand Sketching

Segmentation of Online Sketching Using Geometric Feature

Online Sketch Recognition Using Contexts

A Structural Method for Online Sketched Symbol Recognition.

Coupling Deep Textural and Shape Features for Sketch Recognition.

An Efficient Primitive-Based Method to Recognize Online Sketched Symbols with Autocompletion

Segmentation of Online Freehand Sketching Based on Speed Feature

A recognition algorithm for online sketched symbols with continuous feedback

Online Diagramming Recognition Based on Automatic Stroke Parsing and Bayesian Classifier

On-Line Sketchy Graphic Recognition Based on Adaptive HMM

Sketch Recognition With Natural Correction And Editing

Two-Dimensional Sketch Recognition with Design Intent Capture

A Self-Adaptive Algorithm on Continuous Sketching Recognition