# Multi-Task Tennis Stroke Biomechanics Analysis Using MediaPipe Pose

arXiv: https://arxiv.org/abs/2606.15992 · HTML: https://mlsanity.com/paper/2606.15992/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 10 references that are in the map.
- First posted on arXiv 14 Jun 2026.

## Builds on (most cited first, 4 of 4)

- [A Survey on Multi-Task Learning](https://mlsanity.com/paper/1707.08114/) (2017)
- [MediaPipe: A Framework for Building Perception Pipelines](https://mlsanity.com/paper/1906.08172/) (2019)
- [Action Quality Assessment Across Multiple Actions](https://mlsanity.com/paper/1812.06367/) (2018)
- [TrackNet: A Deep Learning Network for Tracking High-speed and Tiny Objects in Sports Applications](https://mlsanity.com/paper/1907.03698/) (2019)

## Abstract (arXiv)

We built a multi-task pipeline for tennis stroke biomechanics from plain RGB video. On top of pose-based stroke recognition, it adds two new tasks, predicting shot direction and grading posture quality, plus a rule-based feedback layer that suggests coaching tips. Strokes are found automatically using a weighted joint velocity score, s(t) = 0.5 v_wrist + 0.3 m_elbow + 0.2 m_shoulder, removing the need for manual annotation. Pose comes from MediaPipe Pose Landmarker (33 landmarks, metric world coordinates), with each stroke turned into a 30-frame by 39-feature sequence for TennisTransformerGPU, a compact 564,103-parameter transformer (4 layers, 4 heads, d=128) with three parallel output heads. Trained on 1,281 labeled strokes from 7 pros and 1 amateur across 11 videos, it hits 83.7% stroke-type accuracy, 61.9% on direction, and 62.6% on posture under a random 80/20 split. The interesting test is cross-player: train on pros, evaluate on the amateur. Stroke type barely budges, 82.9%, a 0.8% drop. Direction prediction does not transfer; it just falls back to the majority class. An ablation shows why world coordinates matter so much here: switching to image-space landmarks tanks cross-player stroke-type accuracy from 83% to 47% and direction from 68% to 21%. Everything runs on Kaggle's free T4 GPU tier and is fully reproducible.