PROPOSED PROJECT
Sponsor Exposure Tracker
Help clubs and event organizers document sponsor visibility in match footage. Detect supplied logos on jerseys, signage, and interview backdrops, then produce a timeline of appearances, estimated visible seconds, and supporting clips for the sponsorship team to check and include in renewal reports.
North, in view.
60-second review window
Logo match / courtside boardWithin this 60-second example.
- Accepted intervals
- 3
- Excluded replay
- 1
- Overlaps
- Counted once
Put the footage
behind the figure.
North is a fictional sponsor on a basketball court’s advertising board. In a 60-second example excerpt, a reviewer accepts courtside intervals from 04–16 seconds and 28–43 seconds, plus a jersey appearance from 10–18 seconds. The overlap is counted once at sponsor level. A replay from 48–54 seconds is excluded under this report’s rules.
WHAT IS USUALLY DONE
A sponsorship coordinator scrubs through match footage, notes when a logo appears, and takes screenshots for the renewal deck. Replays and overlapping placements can be counted inconsistently, while a page of screenshots gives the sponsor no easy way to check the total.
WHAT SOVRUN CAN DO FOR YOU
Sovrun would build a footage review tool for a club or event organizer. Start with a supplied logo set and one broadcast recording. Computer vision finds candidate logo appearances; the tool groups them into time intervals and shows the matching frames. A reviewer accepts or corrects each interval. The final report links visible seconds to the exact source footage and separates replays, unclear appearances, and placements.
Two placements.
One honest total.
When the sign and jersey are visible at the same time, adding their durations would inflate the sponsor-level result. Keep placement detail, then merge accepted time intervals for the total.
- Board: 12s + 15s
- 27s
- Jersey: 18s − 10s
- 8s
- Overlap: 16s − 10s
- 6s
14 seconds from 04–18, plus 15 seconds from 28–43. The six-second replay stays excluded.
Was the sponsor visible?
The report documents the accepted screen time and where it occurred. It does not measure whether viewers noticed the logo, how many unique people watched, or what the sponsorship was worth.
A report they
can check for themselves.
The first project would cover one match recording and an agreed sponsor list, with a review tool, checked calculations, and evidence that can be included in a renewal discussion.
29 seconds
Accepted visibility in a 60-second excerpt. Replays excluded; overlaps merged.
Open the exact moment.
- Board
- 04–16s
- Board
- 28–43s
- Jersey
- 10–18s
Replay excluded.
- Source interval
- 48–54s
- Reason
- Replay
- Counted time
- 0s
- Rule
- Live play only
How we’d make it happen.
Define what counts
Agree on minimum logo visibility, replay handling, and the reporting unit. Check the recording’s timecode and test a short section against a reviewer’s own notes.
Find and review appearances
Detect candidate matches and group them by placement. Show the source frame, interval, and uncertainty so the team can correct false matches and boundaries.
Report from accepted intervals
Merge overlapping intervals for sponsor-level totals. Retain placement detail and excluded findings, then export the report with source links and sharing permissions.
Where AI helps
Vision models find candidate logos even when their size and angle change. We test on small signs, motion blur, partial cover, and similar marks from the supplied footage. Tracking connects nearby detections, while ordinary time calculations merge overlapping accepted intervals. Uncertain detections go to review rather than into the total.
What people approve
The sponsorship team approves logo matches, interval boundaries, placement labels, and whether replays count for the agreed report. Visible time does not establish audience attention, unique reach, or sponsorship value. Usage permissions determine which supporting clips can be shared.
What we’d need to start
- Match recording with a stable file timecode and permission to analyze it
- Sponsor logo files, placement list, and agreed counting rules
- Reporting template, review owner, and rules for sharing supporting footage
How we’d test it
Compare detections with human-labeled footage from different camera angles. Count missed and false appearances, check interval boundaries, and verify that overlapping placements do not inflate sponsor-level totals. Re-run the same file to confirm stable calculations. Report measured detection quality from that sample, not a promised accuracy rate.
Prepared for Sports & fitness. We would agree on the work after reviewing your brief, materials, and requirements.
This is a proposed project, not completed client work. The visuals and numbers show example outputs, not measured results.
