Dataset Information

Scaling up SoccerNet with multi-view spatial localization and re-identification.

ABSTRACT: Soccer videos are a rich playground for computer vision, involving many elements, such as players, lines, and specific objects. Hence, to capture the richness of this sport and allow for fine automated analyses, we release SoccerNet-v3, a major extension of the SoccerNet dataset, providing a wide variety of spatial annotations and cross-view correspondences. SoccerNet's broadcast videos contain replays of important actions, allowing us to retrieve a same action from different viewpoints. We annotate those live and replay action frames showing same moments with exhaustive local information. Specifically, we label lines, goal parts, players, referees, teams, salient objects, jersey numbers, and we establish player correspondences between the views. This yields 1,324,732 annotations on 33,986 soccer images, making SoccerNet-v3 the largest dataset for multi-view soccer analysis. Derived tasks may benefit from these annotations, like camera calibration, player localization, team discrimination and multi-view re-identification, which can further sustain practical applications in augmented reality and soccer analytics. Finally, we provide Python codes to easily download our data and access our annotations.

SUBMITTER: Cioppa A

PROVIDER: S-EPMC9210334 | biostudies-literature | 2022 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Scaling up SoccerNet with multi-view spatial localization and re-identification.

Cioppa Anthony A Deliège Adrien A Giancola Silvio S Ghanem Bernard B Van Droogenbroeck Marc M

Scientific data 20220621 1

Soccer videos are a rich playground for computer vision, involving many elements, such as players, lines, and specific objects. Hence, to capture the richness of this sport and allow for fine automated analyses, we release SoccerNet-v3, a major extension of the SoccerNet dataset, providing a wide variety of spatial annotations and cross-view correspondences. SoccerNet's broadcast videos contain replays of important actions, allowing us to retrieve a same action from different viewpoints. We anno ...[more]

PMID: 35729183

Similar Datasets

Project description:BackgroundWhile known efficacious preventive health interventions exist, the current capacity to scale up these interventions is limited. In recent years, much attention has focussed on developing frameworks and methods for scale-up yet, in practice, the pathway for scale-up is seldom linear and may be highly dependent on contextual circumstances. Few studies have examined the process of scaling up from decision to implementation nor examined the sustainability of scaled-up interventions. This study explores decision-makers' perceptions from real-world scaled-up case studies to examine how scale-up decisions were made and describe enablers of successful scale-up and sustainability.MethodsThis qualitative study included 29 interviews conducted with purposively sampled key Australian policy-makers, practitioners and researchers experienced in scale-up. Semi-structured interview questions obtained information regarding case studies of scaled-up interventions. The Framework Analysis method was used as the primary method of analysis of the interview data to inductively generate common and divergent themes within qualitative data across cases.ResultsA total of 31 case studies of public health interventions were described by interview respondents based on their experiences. According to the interviewees' perceptions, decisions to scale up commonly occurred either opportunistically, when funding became available, or when a deliberate decision was made and funding allocated. The latter scenario was more common when the intervention aligned with specific political or strategic goals. Decisions to scale up were driven by a variety of key actors such as politicians, senior policy-makers and practitioners in the health system. Drivers of a successful scale-up process included good governance, clear leadership, and adequate resourcing and expertise. Establishing accountability structures and appropriate engagement mechanisms to encourage the uptake of interventions were also key enablers. Sustainability was influenced by evidence of impact as well as good acceptability among the general or target population.ConclusionsMuch like Kingdon's Multiple Streams Theory of 'policy windows', there is a conceptually similar 'window for scale-up', driven by a complex interplay of factors such as political need, strategic context, funding and key actors. Researchers and policy-makers need to consider scalability from the outset and prepare for when the window for scale-up opens. Decision-makers need to provide longer term funding for scale-up to facilitate longer term sustainability and build on the resources already invested for the scale-up process.

Dataset Information

Scaling up SoccerNet with multi-view spatial localization and re-identification.

Publications

Scaling up SoccerNet with multi-view spatial localization and re-identification.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets