Multi-Graph Transformer for Free-Hand Sketch Recognition

TL;DR

This paper introduces a novel Multi-Graph Transformer model that represents sketches as multiple graphs, capturing geometric and temporal features, leading to improved recognition accuracy over existing CNN and RNN methods.

Contribution

It is the first to represent sketches as graphs and apply GNNs for recognition, combining global, local, and temporal information in a unified framework.

Findings

Achieves 72.80% accuracy on Google QuickDraw sketches.

Outperforms all RNN-based models significantly.

Close to CNN-based performance upper bound.

Abstract

Learning meaningful representations of free-hand sketches remains a challenging task given the signal sparsity and the high-level abstraction of sketches. Existing techniques have focused on exploiting either the static nature of sketches with Convolutional Neural Networks (CNNs) or the temporal sequential property with Recurrent Neural Networks (RNNs). In this work, we propose a new representation of sketches as multiple sparsely connected graphs. We design a novel Graph Neural Network (GNN), the Multi-Graph Transformer (MGT), for learning representations of sketches from multiple graphs which simultaneously capture global and local geometric stroke structures, as well as temporal information. We report extensive numerical experiments on a sketch recognition task to demonstrate the performance of the proposed approach. Particularly, MGT applied on 414k sketches from Google QuickDraw: (i) achieves small recognition gap to the CNN-based performance upper bound (72.80% vs. 74.22%), and (ii) outperforms all RNN-based models by a significant margin. To the best of our knowledge, this is the first work proposing to represent sketches as graphs and apply GNNs for sketch recognition. Code and trained models are available at https://github.com/PengBoXiangShang/multigraph_transformer.