|
Photo Searching Based On Its Contents - This is probably, an
area in search technologies, which is still in its infancy
stage. Though AI and computer vision have touched milestone, we
still don't have a commercial technology or platform (though
Riya.com is emerging to fill the void with its face-recognition
and text-recognition online service) which can effectively
address this shortcoming. We generally perform search through
our digital photo collections/library by querying in for
filename and the search engine crawl through our hard drive
looking for that particular filename or pattern (ignoring what
is inside the photo). Even better, is tagging a particular
photo with keywords and then searching over these keywords. But
still, we are left with a search result, which return a set of
photos, possibly containing, the right photo, provided the photo
was named or tagged meaningfully with the photo's inside
geography. This technique seems rather effective for a small
batch of photos. But what if you have thousands of photos? Will
one simply spend hours, days, weeks to name or tag their photo
collection individually. Simply put together, only we, humans,
have this judgmental power to come up with a name or keyword
that relates to the contents/purpose of the photo. It is this
limitation which led my focus and interest to this area of
searching. Though we are years away from having a mature
technology, there is still a semi-automatic workaround solution
which can effectively ease searching through huge collections of
photos. This solution is what I am currently working on and the
result of this effort should be reflected in my upcoming
project, 'Oriens Organizer'.
-------------- IN DEVELOPMENT--------------
Related Links:
|
