SoccerNet-v2: A Dataset and Benchmarks for Holistic Understanding of Broadcast Soccer Videos

Deliège, Adrien; Cioppa, Anthony; Giancola, Silvio; Seikavandi, Meisam Jamshidi; Dueholm, Jacob V.; Nasrollahi, Kamal; Ghanem, Bernard; Moeslund, Thomas B.; Droogenbroeck, Marc Van

doi:10.48550/arxiv.2011.13367

Cited by 3 publications

(10 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They proposed several tasks, ranging from object detection, action recognition, temporal action localization and replay segmentation. Lastly, SoccerNet-v2 [10] extended SoccerNet [19] with more than 300k extra annotations and propose novel tasks that would support the automatic production of soccer broadcast. In particular, SoccerNet-v2 [10] extended the task of action spotting to 17 classes to understand the finegrained details of a soccer game.…”

Section: Related Workmentioning

confidence: 99%

“…Lastly, SoccerNet-v2 [10] extended SoccerNet [19] with more than 300k extra annotations and propose novel tasks that would support the automatic production of soccer broadcast. In particular, SoccerNet-v2 [10] extended the task of action spotting to 17 classes to understand the finegrained details of a soccer game. They also introduced two novel tasks: camera shot segmentation for broadcast editing purposes and replay grounding for highlight and summarization purposes.…”

Section: Related Workmentioning

confidence: 99%

“…They also introduced two novel tasks: camera shot segmentation for broadcast editing purposes and replay grounding for highlight and summarization purposes. In this work, we leverage the finegrained annotations from SoccerNet-v2 [10] and compete in the task of action spotting.…”

Section: Related Workmentioning

confidence: 99%

“…State-of-the-art comparison. We report the results of NetVLAD++ for action spotting (Average-mAP %) on SoccerNet-v2 [10]. We report the performances of our best model over 5 runs and detail its performances for each action class.…”

Section: Video Encodingmentioning

confidence: 99%

“…We run the sliding window of time T along unseen videos with a temporal stride of 1 to report the classaware actionness scores in time. Similar to [19,7,10], we use a Non Maximum Suppression (NMS) module to reduce positive spots closer than a given temporal threshold T N M S .…”

Section: Video Encodingmentioning

confidence: 99%

See 4 more Smart Citations

Temporally-Aware Feature Pooling for Action Spotting in Soccer Broadcasts

Giancola¹,

Ghanem²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Toward the goal of automatic production for sports broadcasts, a paramount task consists in understanding the high-level semantic information of the game in play. For instance, recognizing and localizing the main actions of the game would allow producers to adapt and automatize the broadcast production, focusing on the important details of the game and maximizing the spectator engagement. In this paper, we focus our analysis on action spotting in soccer broadcast, which consists in temporally localizing the main actions in a soccer game. To that end, we propose a novel feature pooling method based on NetVLAD, dubbed NetVLAD++, that embeds temporally-aware knowledge. Different from previous pooling methods that consider the temporal context as a single set to pool from, we split the context before and after an action occurs. We argue that considering the contextual information around the action spot as a single entity leads to a sub-optimal learning for the pooling module. With NetVLAD++, we disentangle the context from the past and future frames and learn specific vocabularies of semantics for each subsets, avoiding to blend and blur such vocabulary in time. Injecting such prior knowledge creates more informative pooling modules and more discriminative pooled features, leading into a better understanding of the actions. We train and evaluate our methodology on the recent large-scale dataset SoccerNet-v2, reaching 53.4% Average-mAP for action spotting, a +12.7% improvement w.r.t the current state-of-the-art.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Video Encodingmentioning

confidence: 99%

Section: Video Encodingmentioning

confidence: 99%

See 3 more Smart Citations

Temporally-Aware Feature Pooling for Action Spotting in Soccer Broadcasts

Giancola¹,

Ghanem²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

PlayNet: real-time handball play classification with Kalman embeddings and neural networks

Mures,

Taibo,

Padrón

et al. 2023

Vis Comput

View full text Add to dashboard Cite

Real-time play recognition and classification algorithms are crucial for automating video production and live broadcasts of sporting events. However, current methods relying on human pose estimation and deep neural networks introduce high latency on commodity hardware, limiting their usability in low-cost real-time applications. We present PlayNet, a novel approach to real-time handball play classification. Our method is based on Kalman embeddings, a new low-dimensional representation for game states that enables efficient operation on commodity hardware and customized camera layouts. Firstly, we leverage Kalman filtering to detect and track the main agents in the playing field, allowing us to represent them in a single normalized coordinate space. Secondly, we utilize a neural network trained in nonlinear dimensionality reduction through fuzzy topological data structure analysis. As a result, PlayNet achieves real-time play classification with under 55 ms of latency on commodity hardware, making it a promising addition to automated live broadcasting and game analysis pipelines.

show abstract

Going for GOAL: A Resource for Grounded Football Commentaries

Suglia¹,

Lopes²,

Bastianelli³

et al. 2022

Preprint

View full text Add to dashboard Cite

Recent video+language datasets cover domains where the interaction is highly structured, such as instructional videos, or where the interaction is scripted, such as TV shows. Both of these properties can lead to spurious cues to be exploited by models rather than learning to ground language. In this paper, we present GrOunded footbAlL commentaries (GOAL), a novel dataset of football (or 'soccer') highlights videos with transcribed live commentaries in English. As the course of a game is unpredictable, so are commentaries, which makes them a unique resource to investigate dynamic language grounding. We also provide state-of-the-art baselines for the following tasks: frame reordering, moment retrieval, live commentary retrieval and play-by-play live commentary generation. Results show that SOTA models perform reasonably well in most tasks. We discuss the implications of these results and suggest new tasks for which GOAL can be used. Our codebase is available at: https://gitlab.com/grounded-sport-convai/goal-baselines.

show abstract

SoccerNet-v2: A Dataset and Benchmarks for Holistic Understanding of Broadcast Soccer Videos

Cited by 3 publications

References 51 publications

Temporally-Aware Feature Pooling for Action Spotting in Soccer Broadcasts

Temporally-Aware Feature Pooling for Action Spotting in Soccer Broadcasts

PlayNet: real-time handball play classification with Kalman embeddings and neural networks

Going for GOAL: A Resource for Grounded Football Commentaries

Contact Info

Product

Resources

About