Taking the challenge

I can’t ignore a good challenge.  Recently, Alfred Thompson challenged me to test out Microsoft’s new CaptionBot application.  He said that he had been having great success with it and challenged me to try it.  The premise is simple; you send it a photo and it describes what it sees.  It’s important to not send personal photos in times like this.

It’s learning so I’ll use my best teacher empathy.  We always try to find the best in our learners, right?

Don’t tell the rest of the Bring IT, Together Committee but I had it open in another window during our meeting last night and was playing around with it so see what I could do with it.

Here are my results…I just dug around some photos from some trips that were on the hard drive and decided to see how they worked.

The Famous Crab

A friend gave me this photo of a crab from a Scuba trip he’d been on.  It was a fond photo for sharing and editing in my Photoshop workshops.  If that was indeed a plate of food, arrangement needs to be revisited!

Niagara Falls

Niagara Falls at night is one of the most spectacular things to witness (and capture with a camera).  I’m thinking the bot needs to go out more!

Philadelphia

Well, if you look past that big bell with the crack in it, there is a man walking with the person with the umbrella in the background.

St. Louis

I guess I was distracted by that large arch thing when I took the photo.  There is indeed a building off to the right. 

St. Louis (again)

This sports fan was fascinated with the chance to take a photo of classic Busch Stadium.  I completely missed the elevated freeway in the background.

San Antonio

I’ve been to San Antonio twice and never fail to be humbled by the Alamo Shrine which served as a mission.

Phoenix

Bingo!

Well, that was fun.  I don’t think I’m ready to start not tagging my own photos anytime soon though.

Have you tried out the Caption Bot with your own pictures?  What kind of success did you have? 

I’m sure that this student will get better over time and learns.  We just need a bit of patience.

2 thoughts on “Taking the challenge

  1. Andrew Forgrave says:

    Interesting. Is the Caption Bot going purely on an attempt at recognizing elements in photos?

    I wonder if it would have more success if it did reverse image searches (think TinEye) and scraped the comments/meta-data from the search results to look for common words or phrases to inform its response.

    Like

  2. this is fascinating. I’m really interested in what it “sees” that we don’t. I wonder what makes it filter out certain things. I seems to focus strongly on people in images, rather than objects. Hmmm.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s