Soccer Event Detection

Automatically detects events/actions of interest in a soccer match

Target users and customers

All content providers may be interested by the soccer event detection technology to manage efficiently their content database. When repurposing applications are targeted, these same providers will see an interest in having this module to help automatically detect events of interest.

The soccer event detection module may also be used by end users to manage their own content at home.

Application sectors

Content structuring & browsing: Knowing where actions of interest are in a soccer match allows structuring of soccer matches. It also gives direct access to events, thus enabling a non linear browsing in the document.
Content retrieval: By adding additional metadata to contents (ie. timecodes of events), the module finds a direct application in the context of content retrieval in video databases.
Content repurposing: Knowing where the events are in a soccer match allows to repurpose this content for other broadcasting channels such as the internet or portable devices. In the later case, building summary from the detected events is of interest for sending this new repurposed content to mobile phones.

Description:

The soccer detection module is a system that detects automatically events/actions of interest in soccer matches. Provided with a set of pre-computed features from a video stream, the system will output a file with the corresponding event timecodes.

The chosen pre-computed features were proved to be discriminative for the task of soccer event detection. They are used as an input for a classification sub-system based on Bayesian networks. This sub-system was trained and its parameters learned offline on a soccer matches database.

Contrary to what exists in literature, the Bayesian model structure was also automatically learned, without using any expert knowledge. The structure learning property enables one to automatically train another model on another type of videos to detect some other type of events, without manually building the structure of the model.

Technical requirements:

The SoccerEventDetection module takes a set of precomputed features from a video stream as input.
Currently running under matlab.

Conditions for access and use:

Module currently available as a matlab prototype only.
Corresponding deliverables are all stated QI – i.e. this module is only available to Quaero partners, on their request. Related IPL is proprietary of Technicolor.