Fast Interactive Object Annotation with Curve-GCN

TL;DR

This paper introduces Curve-GCN, a graph convolutional network that predicts all polygon vertices simultaneously for faster and more efficient object annotation, outperforming previous methods in both automatic and interactive modes.

Contribution

The paper presents a novel end-to-end framework using GCN for simultaneous vertex prediction, improving annotation speed and accuracy over prior sequential models.

Findings

Curve-GCN outperforms Polygon-RNN++ in efficiency and accuracy.

The model runs at 29.3ms in automatic mode and 2.6ms interactively.

Curve-GCN is significantly faster and more accurate than existing approaches.

Abstract

Manually labeling objects by tracing their boundaries is a laborious process. In Polygon-RNN++ the authors proposed Polygon-RNN that produces polygonal annotations in a recurrent manner using a CNN-RNN architecture, allowing interactive correction via humans-in-the-loop. We propose a new framework that alleviates the sequential nature of Polygon-RNN, by predicting all vertices simultaneously using a Graph Convolutional Network (GCN). Our model is trained end-to-end. It supports object annotation by either polygons or splines, facilitating labeling efficiency for both line-based and curved objects. We show that Curve-GCN outperforms all existing approaches in automatic mode, including the powerful PSP-DeepLab and is significantly more efficient in interactive mode than Polygon-RNN++. Our model runs at 29.3ms in automatic, and 2.6ms in interactive mode, making it 10x and 100x faster than Polygon-RNN++.