honestly the table is not very. As in the previous competition, the camera dash wins. It makes no difference how long you went UKM, you can ride and shoot a video and you are a winner.
So why the mapillary sequence
yDwD-N6MTspSR61LozswHg
is not “green” in the grid??
Hi eneerhut,
As you know one user can make pictures from two/three/four cameras at the same time from one point in the car to different views. Then he can manually/automatically move one from GPS tracks to neibours pedestrian segments. In this case he has double of UKM and double of pictures (one from main road and second from pedestrian). This is bug of counting which there now in the system and all see it on the map.
I suggest divide results of users which use more then one device at the same time by two/three/four (depends from quantyty of cameras he use) or delete the sequences which connected to the wrong roads.
That’s not true actually. The optimal interval for images is 5 images per second, but beyond that there is no advantage. This interval has been chosen as it allows us to derive as much map data as possible from each image and create better 3D reconstructions.
Only unique sequences are receiving points, so again a dashcam would need to be travelling on the right roads to generate a score.
@Sapozhnik this is an unfortunate outcome of inaccurate GPS tracks, but I am not away of anyone abusing this deliberately. It would be a very manual and time consuming process to do so. If you can point to an example of this we can consider options.
Also, multiple camera angles should be encouraged, not discouraged.
But they aren’t, currently, with this point system…
That’s true and something we’re keen to fix, but we don’t want to actively discourage people using four cameras either.
The main addition needed here is for the map matching of each sequence to factor in bearing.
Could you please send me any test email and I forward you examples of incorrect counting. sapozhnik@gmail.com
I agree but also in this case you should encourage owners of 360 cameras. For exapmle, multiply his results to 360. They have less speed, less quantities pictures but more productivity
Agree. Will work on 360º and multi-cam in next iteration of scoring system.
Regarding the examples, I’m interested in examples of deliberately manipulating GPS tracks. There is not much we can do about inaccurate GPS for the time being.
When will you make “Kyiv Central” zone?
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-4
I’m curious since I’m not familiar with these things. Wouldn’t the optimal interval be distance based? So when you’re speaking 5 seconds, at what rate of travel? Is that walking? Thank you.
I am happy to hear that.
It is also the first time I hear that.
I was afraid that two times a second was too expensive.
But on foot I do one per second, unless Mapillary does not mind the double cost for little extra value.
example :
BTW - Thanks for getting this all set up. I’m using it for a trial to see how it works. When is the next one? I’d like to ping some places in town to see if a few folks would be interested. For example, the University of North Florida offers GIS courses. Seems like this would be a good excuse to focus on their wonderful little campus and get it all mapped.
@eneerhut
and
support@mapillary.com - request (6628)
When will you make “Boyarka” (Ukraine) zone?
https://mapillary.github.io/mapillary_greenhouse/ctm/boyarka
UPDATE
fix = TRUE
===== !!! IMAGES ONLY 2017…2018 YEAR !!! =====
===== cheater or bug ??? =====
go to
https://mapillary.github.io/mapillary_greenhouse/global-challenge/q1-2019/
see
kaartcam (Mapillary)
ukm=46 img=95950 score=223490
user as TOP-5 …
NONE IMAGES in From=2018-12-27 To=2019-01-31
…and
go to statistics…
============= UKRAINE =================
https://mapillary.github.io/mapillary_greenhouse/ctm/brovary2019-1
kaartcam | 2150 | 1.3 | 894
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-south/
kaartcam | 1622 | 5.8 | 2438
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-2/
kaartcam | 4240 | 1.4 | 889
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-3/
kaartcam | 9400 | 9.2 | 5193
=============== CHINA ====================
https://mapillary.github.io/mapillary_greenhouse/ctm/hongkong/
kaartcam – 87914 images – 40 km
============== conclusion =====================
kaartcam: cheater or bug ???
FOR HISTORY
https://mapillary.github.io/mapillary_greenhouse/global-challenge/q1-2019/
1
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-4/
2
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-1/
3
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-3/
4
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-2/
5
https://mapillary.github.io/mapillary_greenhouse/ctm/hobart/
6
https://mapillary.github.io/mapillary_greenhouse/ctm/irpin/
7
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-5/
8
https://mapillary.github.io/mapillary_greenhouse/ctm/vilareal/
9
https://mapillary.github.io/mapillary_greenhouse/ctm/jacksonville/
10
https://mapillary.github.io/mapillary_greenhouse/ctm/kharkiv-1/
11
https://mapillary.github.io/mapillary_greenhouse/ctm/kharkiv-2/
12
https://mapillary.github.io/mapillary_greenhouse/ctm/san_maurizio_canavese/
13
https://mapillary.github.io/mapillary_greenhouse/ctm/san_carlo_canavese/
14
https://mapillary.github.io/mapillary_greenhouse/ctm/brovary2019-1/
15
https://mapillary.github.io/mapillary_greenhouse/ctm/melbourne_north/
16
https://mapillary.github.io/mapillary_greenhouse/ctm/uzhgorod-1/
17
https://mapillary.github.io/mapillary_greenhouse/ctm/ivrea/
18
https://mapillary.github.io/mapillary_greenhouse/ctm/boyarka/
19
https://mapillary.github.io/mapillary_greenhouse/ctm/horenychi/
20
https://mapillary.github.io/mapillary_greenhouse/ctm/temecula/
21
https://mapillary.github.io/mapillary_greenhouse/ctm/budapest-2/
22
https://mapillary.github.io/mapillary_greenhouse/ctm/istanbul/
23
https://mapillary.github.io/mapillary_greenhouse/ctm/kovyagi/
24
https://mapillary.github.io/mapillary_greenhouse/ctm/ballerup2019-1/
25
https://mapillary.github.io/mapillary_greenhouse/ctm/barnaul/
26
https://mapillary.github.io/mapillary_greenhouse/ctm/kyiv-6/
27
https://mapillary.github.io/mapillary_greenhouse/ctm/guyandotte/
28
https://mapillary.github.io/mapillary_greenhouse/ctm/vigodifassa/
28
https://mapillary.github.io/mapillary_greenhouse/ctm/biella/
30
https://mapillary.github.io/mapillary_greenhouse/ctm/castellon/
31
https://mapillary.github.io/mapillary_greenhouse/ctm/sevastopol/
32
https://mapillary.github.io/mapillary_greenhouse/ctm/valli_di_lanzo/
33
https://mapillary.github.io/mapillary_greenhouse/ctm/la_trinidad-2/
34
https://mapillary.github.io/mapillary_greenhouse/ctm/zusmarshausen/
35
https://mapillary.github.io/mapillary_greenhouse/ctm/hongkong/
36
https://mapillary.github.io/mapillary_greenhouse/ctm/dunkerque/
37
https://mapillary.github.io/mapillary_greenhouse/ctm/stonetown/
38
https://mapillary.github.io/mapillary_greenhouse/ctm/fortaleza/
39
https://mapillary.github.io/mapillary_greenhouse/ctm/hungerford-2/
40
https://mapillary.github.io/mapillary_greenhouse/ctm/leipzig/
41
https://mapillary.github.io/mapillary_greenhouse/ctm/lisboa-2/
42
https://mapillary.github.io/mapillary_greenhouse/ctm/secunderabad/
43
https://mapillary.github.io/mapillary_greenhouse/ctm/uherske_hradiste/
SCORE JSON
Photo report on the receipt of the prize for the first place. Thank you all and see you in the new challenges.
p.s. I promise not to take away the top prizes anymore, probably …
That is not a professional smartphone holder.
selfie stick nonsense
Thanks for the follow-up, @westbam. I hope you have fun with your new toys. We’ll see how things go in the future.