Image classification in instant messengers using CNN

Internet social networks have now become a pervasive application allowing people to easily share text, pictures, and video and audio files. Popular networks include WhatsApp, Facebook, Reddit and LinkedIn. The Wall Street Journal reports that one in three smart phones runs out of space each day so quickly because of too many wishing messages specially good morning messages exchanged between the people. All those early morning well wishes have driven a tenfold increase in the number of Google searches for "Good Morning images" and finding relevant photos in phone gallery becomes a whole new challenge because of all these quippy, upbeat, inspirational wishing(ex: Good Morning, Happy birthday etc.) messages hogging up precious space. In order to weed out all wishing messages they need to be categorized so that all of them can be deleted at a time.Most of the wishing messages are exchanged in the form of images. Identification of these images is difficult as these messages are sent with different images like sun-dappled flowers, adorable toddlers, birds and sunset etc. It would take a lot of time deleting these messages and sorting out the good ones.

Autorentext

I Lakhmi Devi Assistant professor is interested in machine learning.I have tried working o image classification in instant messengers to reduce the memory wastage using CNN(Convolution Neural Network).