Brief summary of my completed and ongoing personal projects

Novel computer vision approach for scale specific generative stick figure as synthetic tribal art works.

A detailed methodology of generation of figures of Warli art (an ancient tribal art) using precise computational geometry, advanced trigonometry and intelligent computer vision techniques, has been illustrated in this paper. The figures have been generated for both the genders- male and female. The visual variances between the genders are attributed to carefully chosen phenotypic traits and differences in anatomical structures, that have been incorporated during generation of the images. A total of 60,000 images have been generated in multiple batches, with both genders in almost equal proportions.

Visual perception for combat intrusion detection and intelligent premonition framework

A detailed methodology of object detection in a smart city setting has been illustrated in this chapter. The presented methodology focuses on intelligent uses of machine learning and deep learning algorithms for effective extraction of desired ROI from a challenging backdrop. This chapter encompasses a Convolutional Neural Network (CNN) based architecture and a Faster RCNN based approach and holistic comparisons are being made between the results obtained from the different approaches.

A deep neural approach towards staining and tinting of monochrome images

The paper presents a system that colourises monochrome images driven by deep-neural-networks without any additional supportive aid. Throughout the paper we explore feature engineering, transfer learning, colour spaces and colour channels and various architectural neural frameworks. Eventually, stained and tinted colorized image or image sequence(video) has been derived from monochrome images, that are more aesthetically pleasing, information rich and more vivid with higher trace ability. The experimental results, illustrating the viability of our methodology and divulging favourable route for future work.