Monday, June 16, 2008

A car and a person?








Surprisingly the blob on the right is identical to the one on the left after a 90deg rotation.

In the absence of enough information of object's identity one search for contextual evidence to force fitting categorization with respect to the world regularities.

As we seen before in this blog, the contextual cognitive module might be unique and it acts as the same for every human task. This is an image processing example, but it could be a natural language processing example.

Even when objects can be identified via intrinsic information, context can simplify the object discrimination by cutting down on the number of object categories, scales and positions that need to be considered.

http://people.csail.mit.edu/torralba/IJCVobj.pdf

1 comments:

Anonymous said...

Interesting... Context can make this blob look like a car or a person,

but it cant make it look like a Ferrari or a Brad Pitt?