Reconstruction or Approximation of environment from edge data. Using pre-trained models, we are analyzing the incoming edge data (CCTV camera footage) and extracting the detected objects and person's pose. These metadata are sent to remote server, where using these metadata, the video scene is reconstructed or approximated on a Real-time manner.