1
00:00:00,140 --> 00:00:07,430
If you have got a picture of a place and you would like to find where this picture is, then you can

2
00:00:07,430 --> 00:00:09,080
use GSP.

3
00:00:09,680 --> 00:00:19,190
GSP is an AI powered geolocation service to find a picture location, and the way this works is that

4
00:00:19,190 --> 00:00:21,890
it always examines the landmarks.

5
00:00:21,890 --> 00:00:27,410
For example, famous buildings in the background, bridges or natural formations.

6
00:00:28,220 --> 00:00:34,610
It also looks for the types of the trees that are in the picture that might only be found in certain

7
00:00:34,610 --> 00:00:35,450
areas.

8
00:00:36,350 --> 00:00:44,480
In addition to that, GSP examines infrastructures like road signs or building styles or even shapes

9
00:00:44,480 --> 00:00:52,490
of power lines, and it is very helpful to confirm where an image has been taken.

10
00:00:52,490 --> 00:00:57,590
So to use GSP, you want to go to the link that you will find in the resources of this lecture.

11
00:00:57,590 --> 00:00:59,510
And it's currently for free.

12
00:00:59,540 --> 00:01:04,310
And then you want to click here in the middle and upload a certain picture.

13
00:01:04,340 --> 00:01:08,810
I'm going to upload the picture that I have uploaded to Yandex before.

14
00:01:09,010 --> 00:01:11,200
So this is it.

15
00:01:11,980 --> 00:01:15,670
And now it's going to start to analyze the picture.

16
00:01:16,750 --> 00:01:23,080
And if I scroll down we can see that this image is in Spain in a city called Cartagena.

17
00:01:23,650 --> 00:01:29,890
We can also see an explanation on how it were able to identify the location, which is by assigned for

18
00:01:29,920 --> 00:01:36,400
Oscar, and that this is a shipyard in Cartagena, Spain, and that the sign is in Spanish and we can

19
00:01:36,400 --> 00:01:38,350
see mountains in the background.

20
00:01:38,860 --> 00:01:44,170
We will also find some coordinates, so we can copy it and then go to Google Maps.

21
00:01:45,610 --> 00:01:50,920
And then we can paste the coordinates and hit enter and find whether this is right or not.

22
00:01:50,950 --> 00:01:52,960
We can see that this is in Cartagena.

23
00:01:52,990 --> 00:01:58,480
The picture was not taken from the sea obviously, but what I always like to do, because this is actually

24
00:01:58,480 --> 00:01:59,320
incorrect.

25
00:01:59,320 --> 00:02:06,820
What I always like to do is just to say Oscar and then the name of the city, which is Cartagena, like

26
00:02:06,820 --> 00:02:07,570
this.

27
00:02:07,960 --> 00:02:12,910
And we can see that this is the place where the image has been taken.

28
00:02:12,910 --> 00:02:19,740
So I'm going to drop this little person in here to Define whether this is accurate or not.

29
00:02:20,910 --> 00:02:22,680
And here we can see the sign.

30
00:02:22,680 --> 00:02:24,750
So let me come back to the image.

31
00:02:25,500 --> 00:02:27,900
This is the large sign.

32
00:02:30,120 --> 00:02:31,950
And this is the main door.

33
00:02:33,270 --> 00:02:38,010
And as you can see this is where the image has been taken.

34
00:02:38,190 --> 00:02:42,330
So it were able to identify that this was in Cartagena.

35
00:02:42,330 --> 00:02:46,860
But it wasn't able to find where this image or picture has been taken.

36
00:02:46,890 --> 00:02:54,000
But if you go to Google Maps and search for the city and the place that is in the picture, then you

37
00:02:54,000 --> 00:02:59,220
might be able to identify where this image or picture has been taken.

38
00:02:59,250 --> 00:03:04,530
If we scroll down a little bit more, we can find some similar buildings.

39
00:03:04,530 --> 00:03:09,870
But in this case, none of these buildings are related to the place that we are searching for.

40
00:03:10,650 --> 00:03:17,400
Another feature that GSP has is that if you have searched for a person, for example, let me search

41
00:03:17,400 --> 00:03:18,390
for a person.

42
00:03:18,420 --> 00:03:23,040
It can identify from which country this person is.

43
00:03:23,160 --> 00:03:28,990
So we can see That this person is from India, but he is not from Mumbai.

44
00:03:29,590 --> 00:03:35,830
It gives us some random explanation about the architecture and buildings and vegetations, which is

45
00:03:35,830 --> 00:03:37,930
not related to this picture at all.

46
00:03:37,930 --> 00:03:42,430
But somehow it knew that this person is from India.

47
00:03:42,460 --> 00:03:49,930
Let me upload a different picture, for example, this one, and you can see that GSP has identified

48
00:03:49,930 --> 00:03:57,820
that this picture were taken in Nepal and India, and it were taken in India, but not in Nepal.

49
00:03:58,390 --> 00:04:05,170
So sometimes GSP will show you incorrect search results, especially that this temple is blurred.

50
00:04:05,170 --> 00:04:07,270
So it's very hard to identify.

51
00:04:07,270 --> 00:04:11,440
But I would like to cover this in a future section.

52
00:04:11,920 --> 00:04:17,530
So in this lecture you have learned how you can use GSP to identify where a certain picture has been

53
00:04:17,530 --> 00:04:18,250
taken.

54
00:04:18,280 --> 00:04:23,980
Not all search results are correct, so you always need to check whether the information that you have

55
00:04:23,980 --> 00:04:32,080
got are true or not by going to Google Maps, Bing Maps or Yandex Maps, etc. just to validate the accuracy

56
00:04:32,080 --> 00:04:33,490
of the information.