Real-time CNN-based Segmentation Architecture for Ball Detection in a Single View Setup

TL;DR

This paper introduces a real-time CNN-based segmentation method for detecting a ball from a single view, effectively handling challenging conditions like poor contrast and frequent interactions, with improved accuracy through test-time augmentation.

Contribution

It presents a novel CNN architecture that processes image pairs for real-time ball detection and releases a new dataset for this task.

Findings

Real-time inference without temporal delay

Test-time data augmentation improves accuracy

Effective segmentation in challenging conditions

Abstract

This paper considers the task of detecting the ball from a single viewpoint in the challenging but common case where the ball interacts frequently with players while being poorly contrasted with respect to the background. We propose a novel approach by formulating the problem as a segmentation task solved by an efficient CNN architecture. To take advantage of the ball dynamics, the network is fed with a pair of consecutive images. Our inference model can run in real time without the delay induced by a temporal analysis. We also show that test-time data augmentation allows for a significant increase the detection accuracy. As an additional contribution, we publicly release the dataset on which this work is based.