Saturday, August 1, 2015

Making sense of Wi-Fi Sense - Different implications of Home vs. Business!

With the roll out of Windows 10, the feature that is ruffling the most feathers is Wi-Fi sense.

What is Wi-Fi Sense? 

Wi-Fi Sense lets you "automatically" share your password to access Internet using your router with a "friend". Before you get all hot and bothered, there are a few caveats:

  • You have to first tell Windows 10 that you do want to share that particular router's information
  • You then have to select which group of contacts (outlook, skype or Facebook) will be able to get this information
  • The contacts have to have a Windows 10 device
  • The password that is shared is encrypted
  • The contact cannot share the connection information with their contacts without knowing the actual password - which they don't because it is encrypted.
  • They cannot connect with Wi-Fi Sense unless they themselves as shared one of their own connections.

What does this all mean in practical terms?

Setting 1 - Home

The most likely setting is when you have a friend visiting your home and needs Internet access.

Scenario 1.
If the following four conditions are met:
  1. They have a Windows 10 device
  2. They have WiFi Sense turned on and are sharing one of their own connections
  3. You have WiFi Sense turned on and are sharing you own Internet connection
  4. They are in your or Skype or FB contact lists
Then they will be automatically connected to the Internet without asking you for the password.
They will not be able to share this information with their own contacts (unless those contacts are also your contacts in one of the three groups listed above).
WiFi Sense options in Windows 10

Thus this scenario does not pose much of a concern for me.

Scenario 2.
There is the alternative scenario that is more bothersome.  Suppose you did NOT turn on WiFi Sense.  Then they will have to ask you for the password.  Often you will just tell them the password or write it down for them.  If they have a Windows 10 device, they can now share this information with their contacts.  This is because they just need to know you password when they elect to share the connection information.

So my suggestion:
If a friend has a Windows 10 device, do one of two things
  1. Turn on WiFi Sense so they get an encrypted password.  The conditions listed above in Scenario 1. would still need to be met.  OR a lot simpler:
  2. Offer to type in the password for them (without them seeing it) on their device and select
    'Not shared'.  If they ever want to turn sharing on, they will be asked for the password which they will not have.  

Setting 2 - Small Business

Now take a different setting - a small business for example
They could actually turn WiFi Sense on and tell people to connect with them on Facebook.
If they do, they would automatically get free WiFi!
This might become a great marketing tool once Windows 10 becomes more ubiquitous.
Of course the person would need to have to use their own Internet access to connect to Facebook first and then they would be able to get WiFi through the small business or they could have added them on Facebook ahead of time.
I am not sure if this would work right now for FB pages or just for FB friends.  But with this does seem quite possible in the not too distant future.

Sunday, May 31, 2015

Creating 3D models with your SmartPhone

Creating a 3D digital representation of a real life physical object seemed like an impossible task for an amateur. But now this has become quite simple for anyone with a smartphone.

I tried my hand at this using 123D Catch (by Autodesk)  to explore the ease of use, quality of output and ability to share and export.  

Here is the link

To view the 3D rendering

  1. Click on the image or link above
  2. On the page that loads, click on 3D view (at the bottom of the image)
  3. You can use your mouse for navigation control (click on "?" for instructions or see below)
What did I learn from this exercise?
  1. The process was quite simple - all you need is the app and a smartphone
  2. Make sure you take shots from about the same distance, with good lighting and follow the grid on the app.  It is a lot like photosphere or other similar apps
  3. The upload and processing takes significant amount of time.  About 1-2 hours.
  4. Once done, you can download the files for the model (*stl) 
  5. You can make your model public and share with others.  You can tweet, Pin, etc.  You also get an embed code which for some reason did not work for this blog post.
Next steps:
Explore use of this tool for anatomy education!

Tips for navigation
  1. Zoom in to the drummer - on touch pad you may have to use 2 finger slide, on mouse use scroll wheel
  2. Pan to center the drummer - Shift + left click + Drag (slide on touch pad)
  3. Rotate - Left Click + Drag (slide)

Thursday, May 28, 2015

Google Photos - Machine Learning that is Insanely Good!

Google just released the new version of Google Photos for desktop, Android and iOS and after using it for just a few minutes I am completely blown away.

While it is many great features, the "search for photographs" is what took my breath away.  I started off searching for "Wedding", "Beach" etc and then moved on to specific landmarks like "Taj Mahal" and "Chichentiza" and it performed remarkably well.

Then I tried "cricket" and the results were stunning.
Here is a sample of what it found in my photo library:

1.  This was not a big deal - it had tons of clues that it was a cricket match
This was from a famous test match in England
2.  The next one was a screen capture from a smartphone of a scorecard from a cricket match

This had the word "Cricket" in the photo
 3.  This was when I started to really be amazed.  This was a screen capture I took of 3 women from Afghanistan celebrating a famous victory.  The only hint here was the score at the bottom.

Afghan women celebrating a famous victory
 4.  This one had few background clues to this being a cricket match.  It is possible since I had grabbed this from the web that Google had indexed it as being a game of cricket.
Cricket match in ice and snow.

5.  This one was totally amazing.  The FIFA trophy would make you think that Google would index this under soccer.  But the 2 people flanking the trophy are 2 famous cricket players.  Google recognized them and brought this up in the search for cricket.  
Tendulkar and Ganguly 
 6.  And lastly - this grainy shot of my TV showing a cricket match - all you can see is the helmet.
Cricket match on TV

Between the amazing experience with Skype translator and this one with Google Photos it seems that machine learning has reached beyond our wildest expectations.