Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Player Re-Identification in Basketball

Chen, Zhiren LU and Mu, Ran LU (2026) MAMM15 20261
Department of Design Sciences
Ergonomics and Aerosol Technology
Abstract
Player Re-Identification (Re-ID) is a fundamental component for comprehensive team sports analysis. However, applying traditional 2D appearance-based trackers presents unique challenges, as confined courts, complex tactics, and highly similar team uniforms cause severe mutual occlusions and frequent identity switches. Furthermore, erratic player movements and image degradation severely compromise visual identification, making long-term Re-ID exceptionally difficult. To address these domain-specific challenges, this thesis proposes a robust, multi-modal Re-ID framework. Rather than relying solely on global appearance features, our system extracts three orthogonal modalities: omni-scale visual appearance, explicit semantic jersey numbers via... (More)
Player Re-Identification (Re-ID) is a fundamental component for comprehensive team sports analysis. However, applying traditional 2D appearance-based trackers presents unique challenges, as confined courts, complex tactics, and highly similar team uniforms cause severe mutual occlusions and frequent identity switches. Furthermore, erratic player movements and image degradation severely compromise visual identification, making long-term Re-ID exceptionally difficult. To address these domain-specific challenges, this thesis proposes a robust, multi-modal Re-ID framework. Rather than relying solely on global appearance features, our system extracts three orthogonal modalities: omni-scale visual appearance, explicit semantic jersey numbers via a domain-adapted Scene Text Recognition pipeline (YOLO11 and PARSeq), and fine-grained shoe color attributes using zero-shot segmentation (SAM 3). To mitigate single-frame anomalies caused by dynamic sports movements, we implement a temporal feature aggregation module utilizing average pooling for continuous visual features and cumulative majority voting for discrete semantic identifiers. At the tracking level, these modalities are dynamically fused using an XGBoost-based classification engine to effectively handle varying feature reliability. Additionally, a Dynamic Global ID mechanism employing an Exponential Moving Average (EMA) update strategy smoothly adapts to continuous visual drift while preserving historical robustness. Extensive evaluations on the TrackID3x3 dataset demonstrate that our multi-modal approach significantly outperforms traditional baselines, establishing a highly resilient paradigm for continuous player tracking and complex sports analytics. (Less)
Popular Abstract
Watching a fast-paced basketball match is dizzying, especially when teammates wear identi-
cal uniforms. For artificial intelligence, this is a nightmare challenge. Motion blur, rapid
movements, and frequent occlusions cause standard tracking systems to lose their targets.
To fix this, a new framework has been engineered—moving beyond overall appearance to
dynamically read jersey numbers and even “look at the shoes” to tell players apart.
Please use this url to cite or link to this publication:
author
Chen, Zhiren LU and Mu, Ran LU
supervisor
organization
alternative title
A Multi-Modal Feature Fusion Framework for Player Re-Identification in Basketball Scenarios
course
MAMM15 20261
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Player Re-Identification, Multi-Modal Fusion, Temporal Aggregation, Scene Text Recognition, Sports Analytics
language
English
id
9230231
date added to LUP
2026-06-02 13:48:20
date last changed
2026-06-02 13:48:20
@misc{9230231,
  abstract     = {{Player Re-Identification (Re-ID) is a fundamental component for comprehensive team sports analysis. However, applying traditional 2D appearance-based trackers presents unique challenges, as confined courts, complex tactics, and highly similar team uniforms cause severe mutual occlusions and frequent identity switches. Furthermore, erratic player movements and image degradation severely compromise visual identification, making long-term Re-ID exceptionally difficult. To address these domain-specific challenges, this thesis proposes a robust, multi-modal Re-ID framework. Rather than relying solely on global appearance features, our system extracts three orthogonal modalities: omni-scale visual appearance, explicit semantic jersey numbers via a domain-adapted Scene Text Recognition pipeline (YOLO11 and PARSeq), and fine-grained shoe color attributes using zero-shot segmentation (SAM 3). To mitigate single-frame anomalies caused by dynamic sports movements, we implement a temporal feature aggregation module utilizing average pooling for continuous visual features and cumulative majority voting for discrete semantic identifiers. At the tracking level, these modalities are dynamically fused using an XGBoost-based classification engine to effectively handle varying feature reliability. Additionally, a Dynamic Global ID mechanism employing an Exponential Moving Average (EMA) update strategy smoothly adapts to continuous visual drift while preserving historical robustness. Extensive evaluations on the TrackID3x3 dataset demonstrate that our multi-modal approach significantly outperforms traditional baselines, establishing a highly resilient paradigm for continuous player tracking and complex sports analytics.}},
  author       = {{Chen, Zhiren and Mu, Ran}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Player Re-Identification in Basketball}},
  year         = {{2026}},
}