Research Projects

Government has taken an initiative to preserve the culture and heritage. The objective of this proposal to create a 3D virtual reconstruction of a lost structure excavated from Kheezhadi site. It is essential to build the archaeological evidence and textual evidence of the past which depicts the ancient civilization, culture and tradition followed in the late sangam period. It is a highly challenging to construct the past from the currently available lost structures of the objects. In order to develop a precise 3D virtual models from the broken relics acquired from archaeological site, deep learning network such as Generative Adversarial Network(GAN) is to be adopted. For building this, convolutional neural network is required to estimate the shape from broken and lost pieces which further necessitates large collection of the data. Hence the first phase of this proposal is to collect the data related to archaeological and ethnoarchaeological information from various Museums, sculptures and Monuments. Later the collected data is used to construct 3D virtual objects from the estimated point clouds through Generative adversarial network model. The experiential model is to be compiled and collated with the understanding of history, ancient stories and documentation. Finally, this model is reconstructed as a virtual environment to visualize the scene of Kheezhadi with the concepts and facts