Production automation should consider the widespread adoption of automatic information extraction tools. This originates mainly from a cultural and business perspective. The MIM/SCAIE project intends to propose new tools for producing more information (including metadata) needed by modern production systems, at lower cost. Explaining, enabling and supporting the introduction of these techniques is the major objective of MIM/SCAIE.

Business examples

Archive management applications
Digital Rights Management infrastructures may benefit from the employment of automatic copy detection techniques and editorial versions tracking methods. For archive retrieval, automatic information extraction tools can provide advanced indexing and retrieval techniques based on content analysis and text semantic mining.

Multi-purpose and multi-channel productions
Multi-channel production and delivery platforms are considered the entry-level infrastructures of the near future for broadcasters. To enhance their automation capabilities, information extraction tools may be used to e.g. identify regions of interest in audio/video clips, or to detect relevant objects to be preserved in re-conditioning operations.

News production
News production environments are characterised by strict requirements in terms of content management and access. Advanced indexing techniques can substantially improve the overall efficiency and time-to-market of these productions. Content summarisation techniques for automated highlight clips extraction or video skimming are two examples.

Ongoing work

- New test material has been collected and more work will be conducted on ground truth material.
- Agreement on a mpeg-7 profile, now at working draft level in MPEG

- The Request for Technology (RFT) can be downloaded from here.

- The group will work on a framework for on-demand evaluation of tools

FP7 TOSCA

- A lot of activities related to MIM/SCAIE will take place in the TOSCA project, which counts as members or liaisons many  MIM/SCAIE members.