SportRGDET: Scene prior and relation graph enhanced detection transformer for sports scenes
- 1Department of Physical Education, Xiamen University, Xiamen, Fujian, China
- 2Department of Physical Education, Sun Yat-sen University, Guangzhou, Guangdong, China
- 3Sports Culture Research Center, Tianjin University of Sport, Tianjin, China
Abstract
Object detection in sports videos holds significant application value in tactical analysis, athlete tracking, and intelligent broadcasting. However, existing methods face two key challenges when processing sports scenes: on one hand, target distribution in sports footage is strictly constrained by field boundaries, yet generic detectors uniformly allocate detection resources across the entire image, causing frequent false detections in off-field regions due to the visual similarity between spectators and athletes; on the other hand, objects such as athletes, balls, and referees in sports scenes form stable spatial co-occurrence patterns governed by game rules, yet existing methods predict each object independently, failing to leverage cross-category relational cues to assist the detection of difficult targets. To address these issues, this paper proposes SportRGDET (Sport Relation-Graph DEtection Transformer), a framework that achieves accurate detection through two core modules: the Scene Prior Guided Module (SPGM) jointly estimates the spatial probability distribution and boundary gradient map of the field region from multi-scale encoder features, fuses them to generate a hierarchical field confidence mask, and performs spatial modulation on the deformable self-attention weights of the encoder, which reallocates attention resources toward the effective field area and progressively suppresses off-field interference across layers, thereby reducing false detections while incurring only moderate additional computation; the Relation Reasoning Enhancement Module (RREM) treats each candidate object as a node during the detection decoding stage, adaptively learns inter-node association strengths through a relation encoder based on category embeddings and relative spatial positions of node pairs, constructs a dynamic relation graph, and bidirectionally propagates contextual information across nodes via graph attention mechanisms, enabling difficult targets to refine their predictions by leveraging positional and categorical cues from surrounding identified objects, enhancing detection robustness under occlusion and blur conditions. Experimental results demonstrate that SportRGDET achieves significant performance improvements on multiple sports detection benchmark datasets, validating the effectiveness of scene prior and relation reasoning mechanisms in sports scene object detection.
Related
- CuFormer U-Net: a transformer-enhanced multi-task network for cucumber leaf disease detection and segmentation
- Improved vision transformer for sports image classification via transfer learning and knowledge distillation
- Trusting the AI rather than the expert? Evidence from sports-related assessments in relation to the Olympic Games 2024.
- Hybrid local–global vision transformer framework for robust plant disease detection in precision agriculture
- An adaptive transformer-based model with entropy-driven knowledge distillation for single-channel EEG REM sleep detection
BibTeX
@article{ouyang2026sportrgdet,
title = {{SportRGDET: Scene prior and relation graph enhanced detection transformer for sports scenes}},
author = {Qingya Ouyang and Chaoyi Su and Qingyuan Li},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0348924},
url = {https://doi.org/10.1371/journal.pone.0348924},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - SportRGDET: Scene prior and relation graph enhanced detection transformer for sports scenes
AU - Qingya Ouyang
AU - Chaoyi Su
AU - Qingyuan Li
PY - 2026
DA - 2026/10/16
JO - PLOS One
DO - 10.1371/journal.pone.0348924
UR - https://doi.org/10.1371/journal.pone.0348924
AB - Object detection in sports videos holds significant application value in tactical analysis, athlete tracking, and intelligent broadcasting. However, existing methods face two key challenges when processing sports scenes: on one hand, target distribution in sports footage is strictly constrained by field boundaries, yet generic detectors uniformly allocate detection resources across the entire image, causing frequent false detections in off-field regions due to the visual similarity between spectators and athletes; on the other hand, objects such as athletes, balls, and referees in sports scenes form stable spatial co-occurrence patterns governed by game rules, yet existing methods predict each object independently, failing to leverage cross-category relational cues to assist the detection of difficult targets. To address these issues, this paper proposes SportRGDET (Sport Relation-Graph DEtection Transformer), a framework that achieves accurate detection through two core modules: the Scene Prior Guided Module (SPGM) jointly estimates the spatial probability distribution and boundary gradient map of the field region from multi-scale encoder features, fuses them to generate a hierarchical field confidence mask, and performs spatial modulation on the deformable self-attention weights of the encoder, which reallocates attention resources toward the effective field area and progressively suppresses off-field interference across layers, thereby reducing false detections while incurring only moderate additional computation; the Relation Reasoning Enhancement Module (RREM) treats each candidate object as a node during the detection decoding stage, adaptively learns inter-node association strengths through a relation encoder based on category embeddings and relative spatial positions of node pairs, constructs a dynamic relation graph, and bidirectionally propagates contextual information across nodes via graph attention mechanisms, enabling difficult targets to refine their predictions by leveraging positional and categorical cues from surrounding identified objects, enhancing detection robustness under occlusion and blur conditions. Experimental results demonstrate that SportRGDET achieves significant performance improvements on multiple sports detection benchmark datasets, validating the effectiveness of scene prior and relation reasoning mechanisms in sports scene object detection.
SN - 1932-6203
ER -