Contextual embeddings improve player identification in sports videos
Beyond a Single Crop: Contextualized Embeddings for Player Re-Identification
Computer Vision and Pattern Recognition
Summary
Identifying players in sports videos is tricky because many players on the same team look very similar. The authors propose a new method that not only looks at each player individually but also considers other players in the same scene and across different moments in the video. This approach helps the system better recognize who each player is by using the surrounding players and repeated appearances over time. Their method works better than traditional ways that only look at single images of players.
What this means in practice
- •For sports analytics teams: Improve automatic player tracking and identification by using context from other players and frames in broadcast video.
- •For video surveillance operators: Enhance person recognition in crowded environments by leveraging spatial and temporal context alongside individual appearance.
Authors
Ge Luo, Changhuo Yang, Weisi Fan, Rui Yang, Shaokai Wang
Abstract
Person re-identification is typically addressed by learning identity-discriminative embeddings from individual person crops. However, player re-identification in broadcast sports videos presents a unique challenge: players from the same team often share similar appearances, while video sequences provide repeated observations of the same players under different viewpoints and surrounding contexts. Existing crop-based approaches largely ignore this contextual information, which can be critical for resolving visually similar identities. Here, we propose ContextReID, a contextualized embedding framework that represents a player by jointly considering other players in the same frame and observations from different frames of the broadcast sequence. By incorporating both co-occurring players and cross-frame observations, the model learns identity features relative to the surrounding visual context rather than relying only on isolated appearance cues. We implement this framework with a vision-language model that takes sets of player crops as input and generates context-aware embeddings for individual players. A combined contextual learning and retrieval objective encourages the model to capture cross-frame identity relationships while maintaining discriminative embeddings for player matching. Experiments across three multi-object tracking benchmarks show that ContextReID improves player retrieval over conventional single-crop embedding approaches in both Rank-1 and mAP. These results highlight the utility of spatial and temporal context for reliable player identification in broadcast sports videos.