Semantic Multinomial representation of Images obtained using DNN and Pseudo concept SVMs

Hand-engineered local image features have been proven to be intended representation for a variety of high level visual recognition tasks. But as the visual recognition tasks such as scene classification and object detection become more challenging, the semantic gap between low-level feature and the concept descriptor of the scene images increases. In this work, we are exploring novel semantic multinomial (SMN) image representation that renders it possible to represent natural scenes by complex semantic description.

Segment-Level Pyramid Match Kernels for the classification of varying length patterns of speech using SVMs.

Classification of long duration speech, represented as varying length sets of feature vectors using SVMs requires a suitable kernel. In this work we propose a novel segment-level pyramid match kernel by partitioning the speech signal into increasingly finer segments and matching the corresponding segments.

Scene Image Classification using Spatial Pyramid Match Kernel.

In this work we explore the approaches to classification of scene images for content based image retrieval (CBIR) system. The main focus of this work is using the kernel method based approaches to classification of scene images represented as sets of local feature vectors.

A Content-Aware Image Prior

In this project we explore the image restoration tasks, a heavy-tailed gradient distribution of natural images has been extensively exploited as an image prior. Most image restoration algorithms impose a sparse gradient prior on the whole image. To address this issue we explore an image restoration algorithm that adapts the image prior to the underlying texture.

Vehicle tracking system using automatic number plate recognition.

In this project, two different modules were developed ie., License Plate Extraction which extracts the number plate from the Vehicle Images taken at some traffic light area and then applying Optical Character Recognition by Template Matching Technique using MATLAB.

KEC Learning Portal.

KEC PORTAL aims to provide a live platform for fast exchange of information within college management, faculty and students to carry the overall functioning at various levels of quality and competence.