Multi-Task Tennis Stroke Biomechanics Analysis Using MediaPipe Pose

• URL: http://arxiv.org/abs/2606.15992v1
• Date: Sun, 14 Jun 2026 19:40:45 GMT
• Title: Multi-Task Tennis Stroke Biomechanics Analysis Using MediaPipe Pose
• Abstract summary: Pose comes from MediaPipe Pose Landmarker (33 landmarks, metric world coordinates)<n>Each stroke turned into a 30-frame by 39-feature sequence for TennisTransformer GPU.<n>It hits 83.7% stroke-type accuracy, 61.9% on direction, and 62.6% on posture under a random 80/20 split.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• How Much Velocity Does Off-Ball Space Value Need? A Broadcast-Viewport Benchmark [0.9185901163605424]
  Velocity-aware pitch control is standard, but under a broadcast viewport half the players are off screen.<n>We ask at which layer of broadcast off-ball analysis velocity changes the answer.<n>We score four velocity regimes against a velocity-aware ground truth at three layers: imputation, the control surface, and team verdicts.
  arXiv  Detail & Related papers  (2026-09-09T20:10:24Z)
• Dancing Stick Figures: An Introductory Dataset for Training Video Generation Models [0.0]
  Dancing Stick Figures is a synthetic video dataset built against three obstacles.<n>For speed, its 64x64, 64-frame reference task is sized for practical repeated training on a single workstation GPU.<n>For accessibility, the release is a 0.79-GB training tier of 4,020 video clips--1,340 six-second source motions.
  arXiv  Detail & Related papers  (2026-08-29T08:03:19Z)
• Training-Free Off-Screen Player Imputation for Broadcast-Based Spatial Football Analytics [0.9185901163605424]
  Spatial football metrics assume access to the positions of all 22 players.<n>Broadcast main camera shows only 10-16 of them at any moment.<n>We quantify the resulting distortion with an open, reproducible benchmark.
  arXiv  Detail & Related papers  (2026-07-13T13:33:45Z)
• Efficient Visual Pointing for Embodied AI:Agent-Driven Data Synthesis, Cross-Block Attention, and Iterative Correction [55.11480729304395]
  We describe our PointArena 2026 solution, which achieves 77.2% overall accuracy and ranks second on the benchmark.<n>The ap proach targets three failure modes. First, agent-driven syn thesis builds large semantic and anchor-relative candidate pools.<n>Second, a determinis tic steerable-data pipeline creates a verified 10,000-sample main set, plus reserve samples, using masks, templates, and path verification.
  arXiv  Detail & Related papers  (2026-06-29T06:39:03Z)
• TT4D: A Pipeline and Dataset for Table Tennis 4D Reconstruction From Monocular Videos [20.74832764427194]
  We present TT4D, a large-scale, high-fidelity table tennis dataset.<n>It provides $140+$ hours of reconstructed singles and doubles gameplay from monocular broadcast videos.<n>This rich data provides a new foundation for virtual replay, in-depth player analysis, and robot learning.
  arXiv  Detail & Related papers  (2026-05-02T04:14:40Z)
• Interpretable Pre-Release Baseball Pitch Type Anticipation from Broadcast 3D Kinematics [3.1224081969539714]
  We classify eight pitch types from monocular 3D pose sequences without access to ball-flight data.<n>Our pipeline chains a diffusion-based 3D pose backbone with automatic pitching-event detection.<n>We achieve 80.4% accuracy using body kinematics alone on 119,561 professional pitches.
  arXiv  Detail & Related papers  (2026-03-05T07:04:35Z)
• VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos [62.29924199978745]
  VideoAgentTrek is a scalable pipeline that automatically mines training data from publicly available screen-recorded videos at web scale.<n>Our approach addresses a key challenge: raw videos contain implicit demonstrations but lack explicit action labels.<n>applied to 39,000 YouTube tutorial videos, our pipeline generates 1.52 million interaction steps automatically.
  arXiv  Detail & Related papers  (2025-10-22T11:25:48Z)
• Robot See Robot Do: Imitating Articulated Object Manipulation with Monocular 4D Reconstruction [51.49400490437258]
  This work develops a method for imitating articulated object manipulation from a single monocular RGB human demonstration. We first propose 4D Differentiable Part Models (4D-DPM), a method for recovering 3D part motion from a monocular video. Given this 4D reconstruction, the robot replicates object trajectories by planning bimanual arm motions that induce the demonstrated object part motion. We evaluate 4D-DPM's 3D tracking accuracy on ground truth annotated 3D part trajectories and RSRD's physical execution performance on 9 objects across 10 trials each on a bimanual YuMi robot.
  arXiv  Detail & Related papers  (2024-09-26T17:57:16Z)
• BootsTAP: Bootstrapped Training for Tracking-Any-Point [62.585297341343505]
  Tracking-Any-Point (TAP) can be formalized as an algorithm to track any point on solid surfaces in a video. We show how large-scale, unlabeled, uncurated real-world data can improve a TAP model with minimal architectural changes. We demonstrate state-of-the-art performance on the TAP-Vid benchmark surpassing previous results by a wide margin.
  arXiv  Detail & Related papers  (2024-02-01T18:38:55Z)
• Video-based automatic lameness detection of dairy cows using pose estimation and multiple locomotion traits [0.0]
  The motion of nine keypoints was extracted from videos of walking cows.<n>The trajectories of the keypoints were then used to compute six locomotion traits.<n>The three most important traits were back posture measurement, head bobbing, and tracking distance.
  arXiv  Detail & Related papers  (2024-01-10T14:56:54Z)
• Fine-Grained Action Detection with RGB and Pose Information using Two Stream Convolutional Networks [1.4502611532302039]
  We propose a two-stream network approach for the classification and detection of table tennis strokes. Our method utilizes raw RGB data and pose information computed from MMPose toolbox. We can report an improvement in stroke classification, reaching 87.3% of accuracy, while the detection does not outperform the baseline but still reaches an IoU of 0.349 and mAP of 0.110.
  arXiv  Detail & Related papers  (2023-02-06T13:05:55Z)
• P2ANet: A Dataset and Benchmark for Dense Action Detection from Table Tennis Match Broadcasting Videos [64.57435509822416]
  This work consists of 2,721 video clips collected from the broadcasting videos of professional table tennis matches in World Table Tennis Championships and Olympiads. We formulate two sets of action detection problems -- emphaction localization and emphaction recognition. The results confirm that TheName is still a challenging task and can be used as a special benchmark for dense action detection from videos.
  arXiv  Detail & Related papers  (2022-07-26T08:34:17Z)
• Take an Emotion Walk: Perceiving Emotions from Gaits Using Hierarchical Attention Pooling and Affective Mapping [55.72376663488104]
  We present an autoencoder-based approach to classify perceived human emotions from walking styles obtained from videos or motion-captured data. Given the motion on each joint in the pose at each time step extracted from 3D pose sequences, we hierarchically pool these joint motions in the encoder. We train the decoder to reconstruct the motions per joint per time step in a top-down manner from the latent embeddings.
  arXiv  Detail & Related papers  (2019-11-20T05:04:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.