Widely Applicable Strong Baseline for Sports Ball Detection and Tracking

TL;DR

This paper introduces a versatile and robust baseline method for sports ball detection and tracking that outperforms existing approaches across multiple sports datasets, supported by new datasets and annotations.

Contribution

The paper proposes a new SBDT baseline with high-resolution features, position-aware training, and temporal inference, validated on multiple datasets with extensive comparisons.

Findings

Our method outperforms 6 state-of-the-art SBDT methods across 5 datasets.

New datasets and annotations facilitate comprehensive evaluation.

The approach demonstrates wide applicability across different sports categories.

Abstract

In this work, we present a novel Sports Ball Detection and Tracking (SBDT) method that can be applied to various sports categories. Our approach is composed of (1) high-resolution feature extraction, (2) position-aware model training, and (3) inference considering temporal consistency, all of which are put together as a new SBDT baseline. Besides, to validate the wide-applicability of our approach, we compare our baseline with 6 state-of-the-art SBDT methods on 5 datasets from different sports categories. We achieve this by newly introducing two SBDT datasets, providing new ball annotations for two datasets, and re-implementing all the methods to ease extensive comparison. Experimental results demonstrate that our approach is substantially superior to existing methods on all the sports categories covered by the datasets. We believe our proposed method can play as a Widely Applicable Strong Baseline (WASB) of SBDT, and our datasets and codebase will promote future SBDT research. Datasets and codes are available at https://github.com/nttcom/WASB-SBDT .