Event-based Batting Impact Estimation

Ryotaro Ishida†, Wataru Ikeda†, Ryosei Hara†
Akemi Kobayashi‡, Toshitaka Kimura‡, Mariko Isogawa†
† Keio University    ‡ NTT Communication Science Laboratories
ICIP 2026

Left: Input event data. Center: Raw masks produced by SAM 3 (these may include merged ball and bat masks at impact). Right: Refined masks from our Mask Refinement Network, which accurately extract ball and bat contours and enable precise impact-timing estimation.

Comparison of sensing modalities
Comparison of sensing modalities.
Conventional RGB video at 120 fps suffers from motion blur around impact, making impact timing estimation less accurate, whereas event-camera data at 10,000 fps provides the temporal resolution needed for precise impact timing.

Abstract

Estimating the precise timing of batting impact is crucial for understanding the rapid sensorimotor control. However, this task is challenging for RGB cameras due to insufficient temporal resolution and motion blur. Similarly, Inertial Measurement Units (IMUs) are impractical for actual matches due to sensor intrusiveness and their limited temporal precision. To overcome these limitations, we propose a novel framework leveraging event-based cameras, which offer microsecond resolution and high dynamic range, to estimate impact timing based on the weighted centroid distance between the detected ball and bat. To address the domain gap between event frames and RGB images that degrades segmentation accuracy, we generate high-density event frames. We then introduce a mask refinement network that leverages these frames and bidirectional mask information, optimized using a novel loss function. Experiments on real-world datasets demonstrate that our method achieves superior accuracy under challenging conditions, including low-light environments and severe occlusions, outperforming baselines by reducing the Mean Absolute Error by approximately 63%.

The Overview of Proposed Method

Method overview figure

Qualitative Results

Qualitative results figure

BibTeX

@inproceedings{Ishida2026eventimpact,
  title={{Event-Based Batting Impact Estimation}},
  author={Ryotaro Ishida and Wataru Ikeda and Ryosei Hara and Akemi Kobayashi and Toshitaka Kimura and Mariko Isogawa},
  booktitle={IEEE International Conference on Image Processing (ICIP)},
  year={2026}
}