Abstract

The sculptures of Indian temples is threatened by natural erosion, environmental degradation, human interference and changing practices of heritage management. Recent advances of artificial intelligence (AI) combined with computer vision, machine learning and 3D reconstruction have improved the documentation, analysis and digital restoration of cultural heritage. However, the current AI-based preservation approaches mostly rely on visual and geometric data and rarely incorporate the traditional Indian knowledge systems such as Shilpa Shastra, which define canonical proportions, iconographic features and sculptural principles. This limitation may result in technically correct reconstructions that do not fully represent the cultural, symbolic, and spiritual significance of temple sculptures.
Based on a review of AI-based heritage conservation studies and Shilpa Shastra principles, this paper proposes an AI-Shilpa Shastra Integrated Preservation Framework for Indian temple sculptures. The framework integrates AI techniques, along with computer vision and 3D reconstruction, with indigenous sculptural knowledge related to proportions, iconography, gestures, postures, and symbolic attributes. Its objective to combine ancient rule-book (Shilpa shastra) to identify damage based on cultural meaning and digitally rebuilt missing pieces an help experts make the most authentic choice for fixing temple sculptures. Expert validation by art historians, conservators, and traditional artisans is proposed to evaluate the reliability and cultural authenticity of restoration outcomes. The proposed framework provides a conceptual pathway for integrating modern computational methods with indigenous knowledge to support sustainable and culturally sensitive preservation of Indian temple sculptures.