Culture is an important aspect which binds people with their ancestors, imparts knowledge of the past, and acts as a window for the modern era to capture a glimpse of the past. Preservation and proper documentation of culture has become essential as modernization has led to loss of many cultures, but it has also given us tools to preserve it. Rapid advancement in the field of technology has provided us with useful tool which can be leveraged to document all cultural heritage and pass it to the generations to come. Therefore, the work proposes a model to identify the nuances of various Dravidian architectural styles. Dravidian architecture consists of contributions from various empires that ruled southern India and each have their own intricate style contributing to Dravidian architecture. The proposed model utilizes convolutional neural network to identify the nuances, by learning from various images of temples built by three major empires—Pandya, Chola, and Vijayanagara—and predict which empire built the temple.