E. Lecture Notes

G. National Conferences in India

The descriptions that individuals provide about human faces are frequently verbal anyway like, “blonde haired person” or “person with lengthy face”. The work involves instantly removing such semantic descriptions of faces in the image database and performing query in regards to a particular face using verbal descriptions more effective. We presently are utilizing Pruning of images according to description results in lack of the best images because of mistakes through the automated retrieval system or even the user. Hence we use Bayesian Learning for that query and retrieval part.

According to features such as the spectogram, psd, the LP co-efficients of the music wave file, a combination of gaussians may be used to model the extracted features and therefore recognize the instrument playing the bit of music. The work was transported in MATLAB and it was my final year undergraduate project.

A Conuterpropagation neural network was utilized for face recognition and recognition in 2d images. The neural network was modeled so that just by altering the supervised area of the counter propagation network, face recognition might be done with similar network educated to do face recognition. It was my 3rd year (sixth sem) undergraduate project.