Jump to content
Check out the Spin Axis Podcast! ×
IGNORED

Geography Challenge (sorta)


Note: This thread is 3720 days old. We appreciate that you found this thread instead of starting a new one, but if you plan to post here please make sure it's still relevant. If not, please start a new topic. Thank you!

Recommended Posts

Posted

How good are you at guessing where a photo was taken?

https://www.geoguessr.com/

I thought I got a bit lucky on my first try. Each test is 5 pictures, and I got 3 of them somewhat close (but then again, I was guessing "in my wheelhouse of North America and Europe). Only one ended up being in South America. If they had pics from other areas of the world, I'd likely have fared significantly worse. My guesses in orange, the actuals in black below:

56cf35a14f501_ScreenShot2016-02-25at12.0

 

Anyway, I think it's part of an effort to determine the location of any picture, just from the pixels. Here's an article about it:

https://www.technologyreview.com/s/600889/google-unveils-neural-network-with-superhuman-ability-to-determine-the-location-of-almost/

Quote

ere’s a tricky task. Pick a photograph from the Web at random. Now try to work out where it was taken using only the image itself. If the image shows a famous building or landmark, such as the Eiffel Tower or Niagara Falls, the task is straightforward. But the job becomes significantly harder when the image lacks specific location cues or is taken indoors or shows a pet or food or some other detail.


Nevertheless, humans are surprisingly good at this task. To help, they bring to bear all kinds of knowledge about the world such as the type and language of signs on display, the types of vegetation, architectural styles, the direction of traffic, and so on. Humans spend a lifetime picking up these kinds of geolocation cues.

So it’s easy to think that machines would struggle with this task. And indeed, they have.

Today, that changes thanks to the work of Tobias Weyand, a computer vision specialist at Google, and a couple of pals. These guys have trained a deep-learning machine to work out the location of almost any photo using only the pixels it contains.

...

Anyone can play at www.geoguessr.com. Give it a try—it’s a lot of fun and more tricky than it sounds.

Needless to say, PlaNet trounced the humans. “In total, PlaNet won 28 of the 50 rounds with a median localization error of 1131.7 km, while the median human localization error was 2320.75 km,” say Weyand and co. “[This] small-scale experiment shows that PlaNet reaches superhuman performance at the task of geolocating Street View scenes.”

 

  • Upvote 1

My Swing


Driver: :ping: G30, Irons: :tmade: Burner 2.0, Putter: :cleveland:, Balls: :snell:

Awards, Achievements, and Accolades

Posted

I played twice but forgot to screenshot my results. Both were in the 5000s range thanks to a few answers that were spot on and a few that were cross-continental.

In my bag:

Driver: Titleist TSi3 | 15º 3-Wood: Ping G410 | 17º 2-Hybrid: Ping G410 | 19º 3-Iron: TaylorMade GAPR Lo |4-PW Irons: Nike VR Pro Combo | 54º SW, 60º LW: Titleist Vokey SM8 | Putter: Odyssey Toulon Las Vegas H7

Awards, Achievements, and Accolades

Posted
2 minutes ago, jamo said:

I played twice but forgot to screenshot my results. Both were in the 5000s range thanks to a few answers that were spot on and a few that were cross-continental.

My next two I just tried were 3500 an 5000. Yikes. It also lets you do specific countries, even cities, which is cool.

My Swing


Driver: :ping: G30, Irons: :tmade: Burner 2.0, Putter: :cleveland:, Balls: :snell:

Awards, Achievements, and Accolades

  • Moderator
Posted

What the Fudge?

56cf3e157bfe9_ScreenShot2016-02-25at12.4

Steve

Kill slow play. Allow walking. Reduce ineffective golf instruction. Use environmentally friendly course maintenance.

Awards, Achievements, and Accolades

Posted
2 minutes ago, nevets88 said:

What the Fudge?

56cf3e157bfe9_ScreenShot2016-02-25at12.4

Looks like a Bosnian tree

My Swing


Driver: :ping: G30, Irons: :tmade: Burner 2.0, Putter: :cleveland:, Balls: :snell:

Awards, Achievements, and Accolades

  • Moderator
Posted
3 minutes ago, RandallT said:

Looks like a Bosnian tree

I get that you have to zoom out, rotate, move around. You can check out the road signs, which side the car is on and all that, but still lots of guesses for countries that drive opposite the way the US does and the license plates are shaded out.

Steve

Kill slow play. Allow walking. Reduce ineffective golf instruction. Use environmentally friendly course maintenance.

Awards, Achievements, and Accolades

Posted

I got only 5350. Primarily because I mixed up a few ethnicity with regards to spanish, latino, and brazil.

 

Matt Dougherty, P.E.
 fasdfa dfdsaf 

What's in My Bag
Driver; :pxg: 0311 Gen 5,  3-Wood: 
:titleist: 917h3 ,  Hybrid:  :titleist: 915 2-Hybrid,  Irons: Sub 70 TAIII Fordged
Wedges: :edel: (52, 56, 60),  Putter: :edel:,  Ball: :snell: MTB,  Shoe: :true_linkswear:,  Rangfinder: :leupold:
Bag: :ping:

Awards, Achievements, and Accolades

Posted
10 minutes ago, nevets88 said:

What the Fudge?

56cf3e157bfe9_ScreenShot2016-02-25at12.4

Iowa.

Gambling is illegal at Bushwood sir, and I never slice.   

           

Awards, Achievements, and Accolades

Posted

I just did one in the US. Got nearly 5000 points in one question because I nearly nailed one from Lancaster, PA.

In my bag:

Driver: Titleist TSi3 | 15º 3-Wood: Ping G410 | 17º 2-Hybrid: Ping G410 | 19º 3-Iron: TaylorMade GAPR Lo |4-PW Irons: Nike VR Pro Combo | 54º SW, 60º LW: Titleist Vokey SM8 | Putter: Odyssey Toulon Las Vegas H7

Awards, Achievements, and Accolades

  • Moderator
Posted

Do what you know.

56cf424e6c5f4_ScreenShot2016-02-25at1.04

Steve

Kill slow play. Allow walking. Reduce ineffective golf instruction. Use environmentally friendly course maintenance.

Awards, Achievements, and Accolades

Posted

They need a version that shows a golf hole - and then you have to try to pick the course, or pin your guess on a map

  • Upvote 1

Players play, tough players win!

Awards, Achievements, and Accolades

Posted

These are crazy - I just keep seeing open fields of nearly nothing and play the odds and guess Russia.  I managed to get close-ish on 3/5 the second time I played and tallied 9k points.  That was including mistaking northern Australia for central Africa and getting ZERO for it. :P

Awards, Achievements, and Accolades

  • Moderator
Posted
3 minutes ago, Golfingdad said:

These are crazy - I just keep seeing open fields of nearly nothing and play the odds and guess Russia.  I managed to get close-ish on 3/5 the second time I played and tallied 9k points.  That was including mistaking northern Australia for central Africa and getting ZERO for it. :P

The big cities are way way more fun.

56cf47606d4b9_ScreenShot2016-02-25at1.25

Hmmmmm...

56cf47c7bd747_ScreenShot2016-02-25at1.27

Steve

Kill slow play. Allow walking. Reduce ineffective golf instruction. Use environmentally friendly course maintenance.

Awards, Achievements, and Accolades

Posted

I'm having fun picking a guess and then trying to peg that spot on the map without zooming in.  Easy with places like South Africa or Florida, not as easy with places like Wyoming or Czech Republic. :)

Awards, Achievements, and Accolades

Posted
36 minutes ago, nevets88 said:

Do what you know.

Damn- you've nailed NYC and London.

4 minutes ago, Golfingdad said:

I'm having fun picking a guess and then trying to peg that spot on the map without zooming in.  Easy with places like South Africa or Florida, not as easy with places like Wyoming or Czech Republic. :)

Ha, that's what I did my last two tries- no zooming and super-fast "gut" guesses. I figure it's "close enough" for scoring! Interesting that the computer, which does better than humans, has a median error of over 1000km. Humans, over 2000km median.

Made me think of when we were trying to pinpoint where Jihadi John was executing people. I read articles that experts were doing landscape analysis to pinpoint what the exact location was so that we could target him. His landscape was basically desert, but I remember a few tell-tale ridges or something. I wonder if we nailed him that way. Guess we'll never know.

 

My Swing


Driver: :ping: G30, Irons: :tmade: Burner 2.0, Putter: :cleveland:, Balls: :snell:

Awards, Achievements, and Accolades

  • Moderator
Posted
22 minutes ago, RandallT said:

Damn- you've nailed NYC and London.

Ha, that's what I did my last two tries- no zooming and super-fast "gut" guesses. I figure it's "close enough" for scoring! Interesting that the computer, which does better than humans, has a median error of over 1000km. Humans, over 2000km median.

Made me think of when we were trying to pinpoint where Jihadi John was executing people. I read articles that experts were doing landscape analysis to pinpoint what the exact location was so that we could target him. His landscape was basically desert, but I remember a few tell-tale ridges or something. I wonder if we nailed him that way. Guess we'll never know.

 

I basically knew the general vicinity and then zoomed in and looked at the street signs and landmarks. I enjoy navigating big cities, by foot, car, bicycle, public transportation, etc...

Steve

Kill slow play. Allow walking. Reduce ineffective golf instruction. Use environmentally friendly course maintenance.

Awards, Achievements, and Accolades

Posted

Check it out:

 

image.jpeg

They need a dedicated category of golf courses for us, I believe.  :)  (This one was very easy for me to find when I flipped the camera around, but still impossible for me to nail down after that.  I still missed it my several hundred miles.

 

image.jpeg

Awards, Achievements, and Accolades

Note: This thread is 3720 days old. We appreciate that you found this thread instead of starting a new one, but if you plan to post here please make sure it's still relevant. If not, please start a new topic. Thank you!

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


  • Want to join this community?

    We'd love to have you!

    Sign Up
  • TST Partners

    Carl's Place
    PlayBetter
    Golfer's Journal
    ShotScope
    The Stack System
    FitForGolf
    FlightScope Mevo

    Coupon Codes (save 10-20%): "IACAS" for Mevo/Stack/FitForGolf, "IACASPLUS" for Mevo+/Pro Package, and "THESANDTRAP" for ShotScope. 15% off TourStriker (no code).
  • Posts

    • Day 610 - 2026-06-03 Got some work in between lessons today. Rare late day, teaching until 7:30pm.
    • Let's continue on… Cool. The thing is, nobody's claiming par is "reliable" and par's inclusion piggy-backs in the course rating, which is awfully close to par and, thus, brings par in to make it make sense. Once again, for those in the back… (CR - Par) just makes it really easy to know what kind of score you need to shoot to best, match, or play worse than your handicap index. Yes, when par is different, the players from the higher par tees get an extra stroke (72 vs. 71, the 72s get an extra stroke. That makes sense and is a small complication (more info at https://www.usga.org/content/usga/home-page/handicapping/roh/Content/rules/Committee%20Content/USGA/LG_R6d.htm). However, most of the time, this adjustment will not be needed, as many courses play to the same par for the same genders from all sets of tees. And, the rare times it is needed, par (measured in whole numbers, integers) and strokes (also whole numbers/integers) map easily and the idea is easily grasped. Dean seems to be unaware of the fact that most every golfer carries something orders of magnitude more powerful than the highest end desktop computers available the last time he consulted with the USGA in their pockets. While it is quaint that his club puts printouts by the first tee… get with the times, Dean. Look up your handicap index and course handicap in the GHIN app and get on with it. It's a better system than the one that didn't account — at all — for a difference in the playing conditions (via an algorithm, not a judgment). Dean's assertions about the "less precise system because of par" continues to make absolutely zero sense. Right, it still changed tee to tee. Now it just changes differently… and in a way that more accurately reflects the score you need to shoot to play to your handicap. Previously, a 1.1 index would get 1 stroke on a 66.7/122 par-72 course. Now they give four strokes back to the course and must shoot 68 to play to their handicap. This makes way more sense. The 18-shot difference is a pretty extreme example. Maybe a long course that also offers a par-three set of tees could play that long, but… man, that's not going to be super common. Sensationalistic much, Dean? Also, once those unhappy (complete assumption) golfers realize a) what the change shows them (playing to net par = playing to your index) and b) realizes that their differential is going to be the same… I think they'll get over their initial questions. No. And yet… if he shoots the same scores, he'll get the same handicap index he has now. But he'll know on each course what score he needs to shoot to "play to his handicap." Sheesh, Dean. This stuff isn't that hard to figure out. Enough with the sensationalistic stuff. I don't find it "unacceptable" at all. Then again, I'm not nearly 80 and seemingly incapable of doing basic math these days. No. This literally makes no sense, as that part of the differential calculation and the course handicap calculation remains identical. Good! No. Categorically wrong. They should have been adjusting their handicaps all along. Previously it was by subtracting the course ratings. Which… is still basically what's done, with the addition of the course rating being "baked in" to the course handicap calculation. Dean is wrong here, or doing some math heretofore unknown by the world. When par is the same, what determines the difference in handicaps? The course rating, which Dean loves! Sheesh! You had to things when players were in situations like this before, too. This is getting exhausting. He keeps using words like "less precise" and "unfair" but does not seem to understand what they mean. This is like the Princess Bride meme: "you keep using that word. I do not think it means what you think it means." The caps reduce upward movement. Committees have reign to reduce a player's handicap, and there's still an automatic Exceptional Score Reduction. I'm going through these more quickly now because… well, it's silly how badly Dean misses the mark with this blog post. Dean is literally confusing the upward movement (with the soft and hard caps) here with the exceptional score reduction which is used when lowering handicaps due to an exceptionally good score. The creators of the WHS are handicap experts. They know more about the current state of handicaps/handicapping than the Pope Emeritus. It's been shown to have almost no effect across all handicaps. Yes, some 36s under the old system are now 35s under the new system. Yawn. He should have stopped there. It's easier to apply and makes more sense. This makes no sense. It's "not complex" but players will have to guess? And, for men or women, the stroke index of each hole doesn't change because they play a different set of tees. They get a different number of strokes, but it's always been true that when you get 14 strokes you apply a stroke to stroke index holes 1-14, and when you get 11, to just holes with a SI of 1-11. Objection, your honor. Assumes facts not in evidence. Dean's just out here continuing to make shit up about "the inaccuracy of par" and ignoring that with Par (an integer) came the Course Rating, which he agrees is precise and accurate. No. No, this is inaccurate. Also, as noted, you can randomly assign stroke indexes, and so long as all the low numbers or all the high numbers are not clumped together at the beginning or ends of the 18 holes, matches generally work out the same. This is inaccurate. It is an algorithm that looks at scores. That's it. Also, this is better than a system like the prior one where no such thing existed at all. Wildly inaccurate and off-base. Did they do actual testing? No need. They have millions and millions of rounds and ran many, many, many simulations. That's testing. Dean seems to continue to be unaware of the fact that computers are more powerful now than they were in 2002. But, he's nearly 80, so we can understand if not going so far as to give him a pass on how much he gets wrong. Cool. Noted. For the most part that was because many countries haven't been able to rate enough of their courses. :sigh:
    • Day 3 (3 Jun 26) - More work on keeping arms connected today - hard foam balls with 7i and 5w…..
    • Day 274 6-3 flow drill getting chest through, arms in front. Arms get a little pinned to the side, not as much in front as I want them when I add speed. 
    • Shot 48 yesterday.  For me bogey golf is good.  I was 10 over through 7 and figured with a Par 3 and 4 coming on all I needed was birdie / par to get my 45. I had a great tee shot on #8 and sunk  a 5 footer for birdie, game was coming together, now just needed par on #9. Had a great tee drive and the green was within range for a hoped GIR or nGIR.  But I pulled the shot left into tall weeds and needed to take a drop.  So much for par, but a bogey for 46 is still good for me. I hit my lob wedge to get over a small tree and saw the ball riding nicely  on line to the pin when my club hit the ball a 2nd time on my follow through causing the ball to change directions and ended up @ pin high but along the same tall weeds I just took an unplayable out of.  had no room for a backswing, Just hacked at it and it shot across the green to the rough on the far side.  Needed a chip & 1 putt got a triple bogey. you can see the hole fall apart in the screenshot below.  
×
×
  • Create New...

Important Information

Welcome to TST! Signing up is free, and you'll see fewer ads and can talk with fellow golf enthusiasts! By using TST, you agree to our Terms of Use, our Privacy Policy, and our Guidelines.