[HOT] Launching MapSwipe

antti nissinen anissine at gmail.com
Mon Jul 18 20:31:14 UTC 2016


Hi!
I have noticed the same phenomenon. Especially if you complete the
session and then come back to the same mission there seems to be same
tiles. Maybe they are still under development and recognizing the user
and matching the offered tile sets doesn't work perfectly yet.

Antti

> I might be wrong, but is it possible that I get same tiles more than once?
>
> thx
> Rudy
>
> On 18.07.2016 18:43, Benjamin Herfort wrote:
>>
>> Hey Joost,
>>
>> tiles will be checked by several volunteers. And yes of course it is
>> really interesting to see, how volunteers agree on different tasks.
>>
>> We did some pre-studies using a browser version using pybossa on this
>> that showed that once 4 volunteers agree on 'yes' you can be sure to 98%
>> that this actually is a house. But there is definitely a greater
>> variation for "harder" tasks. Nevertheless also some buildings were
>> missed.
>>
>> Kind regards,
>> Benni
>>
>> --
>> GIScience Research Group Heidelberg University
>> http://uni-heidelberg.de/gis
>>
>> https://www.facebook.com/GIScienceHeidelberg
>> https://twitter.com/GIScienceHD
>>
>>
>>
>> Am 18.07.2016 um 18:27 schrieb joost schouppe:
>>>
>>> Hi,
>>>
>>> Just wanted to take the time to congratulate the team on this. Our tweet
>>> about it got over 20 retweets and 2500 views [1] (which is a lot for an
>>> account with little over 500 followers), and I even saw my boss swiping
>>> away at work (because of my own retweet).
>>> I hope it gets the exposure it deserves, as it's not just fun and useful,
>>> but also improves the sat pic reading skills of our future new mappers.
>>>
>>> I would love to get my hands on the generated data. Do some tiles get
>>> checked twice (and what is the correlation between swipers)? Is there a
>>> higher incidence of "unsure" tiles in some tasks, indicating harder to
>>> read
>>> pictures?
>>>
>>>
>>> [1] https://twitter.com/osm_be/status/753959061796843520
>>>
>>> 2016-07-13 15:57 GMT+02:00 Pete Masters <pedrito1414 at googlemail.com>:
>>>
>>>> Hi all, how are you?
>>>>
>>>> You might have heard mention of a new tool we are planning to launch on
>>>> Friday called MapSwipe. It's a Missing Maps initiative, developed by
>>>> MSF,
>>>> designed to support OSM / HOT mapping. The idea is that people swipe
>>>> through imagery on their phones and identify areas with specified
>>>> features
>>>> (roads, buildings, etc). The resulting data can then be plugged into the
>>>> tasking manager, meaning mappers have less empty bush, jungle, desert to
>>>> scan through, meaning we can all get on with mapping a bit quicker.
>>>>
>>>> This is a first instance funded by MSF, so we will start off with
>>>> Missing
>>>> Maps tasks in there, but we hope to soon open it up to others. The code
>>>> is
>>>> also all on github.
>>>>
>>>> The app will be available on Friday, but there is a landing page already
>>>> ready [1 <http://mapswipe.org/>] and I also did an interview on the BBC
>>>> world service show, Click, last night [2
>>>> <http://www.bbc.co.uk/programmes/p040g4qr>] in case you want to have a
>>>> listen.
>>>>
>>>> When it comes to the launch on Friday, we would love your support in
>>>> getting the word out. The possibilities of using this kind of tool at
>>>> scale
>>>> are fascinating and your help amplifying the message would be superb.
>>>>
>>>> If you have any questions, just give me a shout!
>>>>
>>>> Cheers,
>>>>
>>>> Pete
>>>>
>>>>
>>>>
>>>> [1] http://mapswipe.org/
>>>> [2] http://www.bbc.co.uk/programmes/p040g4qr
>>>>
>>>>
>>>>
>>>> *Pete Masters*
>>>> Missing Maps Project Coordinator
>>>> +44 7921 781 518
>
>
>
> _______________________________________________
> HOT mailing list
> HOT at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot



More information about the HOT mailing list