correction to "Competition dates, prizes and score limits"

[Francesco Potortì]

This message is the same as the previous one I sent with a single correction: the dates of the survey and competition of Track 6, which were reported incorrectly in my previous message.

================================================================
here are the latest news regarding the IPIN competition 2024.

# Dates and times

Dates are in day/month format.

Track 1: 11/10 survey, 13/10 competition
Track 3: 1/10
Track 4: 30/9
Track 6: 10/10 survey, 11/10 competition

Competitors of T3 and T4 are receiving their scoring trial name before the competition starts.  Competitors must start their first scoring trial in the time window 7.00-7.05 UTC.  The second trial for T3 and T4 will start as soon as possible, but no later than 11.00 UTC.


# Prizes

IEEE Dataport (https://ieee-dataport.org/) sponsored the IPIN competition 2024 with up to 8000 USD to distribute among the winners of the four Tracks of the IPIN competition 2024.

Track 1: 2000$ + 1000$  first + second place
Track 3: 1000$ +  500$  first + second place
Track 4: 1000$ +  500$  first + second place
Track 6: 1300$ +  700$  first + second place

Thanks to IEEE Dataport for sponsoring the IPIN competition 2024.


# Limit scores

For each Track, we define a limit score which must not be exceeded by a competitor to be eligible for a prize.  Limit scores are:

Track 1: 10m
Track 3: 15m
Track 4: 30m
Track 6: 40m


# 90th percentile of point error as tie-breaker for close scores

The EvAAL Framework, briefly described at https://competition.ipin-conference.org/, includes a core criterion called "Realistic measurement resolution", which says that scores not farther apart than 50 cm should be considered equal.  In practice, past IPIN competitions have never seen results so close together, with the exception of pre-2021 offsite competitions, when the EvaalAPI interface was not used.

With the continuous improvement of competing systems, we may now see instances where we need a tie-breaker for competitors with similar scores, that is, with similar third quartiles of point errors.  In order to remain faithful to the "Realistic measurement resolution criterion", in 2024 we are using the 90th percentile of point error as tie-breaker.

-- 
Francesco Potortì (ricercatore)        ISTI - CNR, Pisa, Italy		       
Web:    http://fly.isti.cnr.it         Skype:  wnlabisti
Mobile: +39.348.8283.107	       also Telegram, and even Wa