Since I am researching on image captioning and have avid interest in photography I first decided to understand how many types of images are there. There are so many questions that came up from them.

  • Can we understand that this is an image taken from a kaleidoscope or is it a panoramic photo.
  • Is it a sunrise or a sunset?
  • Is clip art image recognition possible and how
  • How to differentiate between past and present photos of same object in the same environment.
  • Can the computer just classify correct what kind an image is ? What category an image belongs to?
  • How to differentiate the vintage print from other duplicates.

When I am reading the papers I think I can do this, this problem looks approachable, but as soon as I start searching for different types of images, boy! I should have known that there are 50 different kinds, I thought of processing what kinds of questions are possible for each of these. These are questions that I will try to answer way into the future because first I need to write them all down first.

The artificial would include comics, clip arts, paintings, sketches (what about real life 3D sketches and that actual 3D object assuming that the sketch is really real-life) I think these are some kind of outliers but it is still worth looking into.

Taking wikipedia as the state of art (there may be something else but let me start with wiki), it lists 80 categories. Here are some of the features of these items that I can think of (there may be many more) So, can a RNN identify all these? Let me go check!

blog comments powered by Disqus


20 January 2016