International Research Journal of Engineering and Technology (IRJET) Volume: 12 Issue: 11 | Nov 2025
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
Artworth AI: Intelligent Valuation Model for Sculptural Artifacts Based on Aesthetic, Historical, and Material Attribute Prof. Shilpa Joshi 1, Rahul2 1Professor, Master of Computer Application, VTU’s CPGS, Kalaburagi, Karnataka, India 2Student, Master of Computer Application, VTU’s CPGS, Kalaburagi, Karnataka, India
----------------------------------------------------------------------------------***-----------------------------------------------------------------------------Abstract- ArtWorth AI employs an intelligent multi-modal 3.OBJECTIVES approach in order to provide accurate valuations of sculptural artefacts based on combining significant A key goal of this research is to create an intelligent characteristics such as aesthetics, historical provenance, Artificial Intelligence System that will allow for the materials and use of ensemble machine learning accurate and transparent valuation of sculpture by technologies along with Natural Language Processing and leveraging Aesthetic, Historical, and Material Data. The Computer Vision to create accurate and transparent pricing system will provide the ability to extract visual and textual predictions with statistically significant confidence characteristics through Computer Vision and Natural intervals. ArtWorth AI creates opportunities for objective, Language Processing, consolidate the physical reliable, and evidence-based sculpture valuation through characteristics into a structured data set, apply an providing insight and detection for Authenticity Risk for ensemble learning approach to predict the price of a both galleries and collectors. sculpture with uncertainty estimation, and offer explainable services for reliable decision-making.
1.INTRODUCTION
4.RESEARCH METHODOLOGY
ArtWorth AI is an Explainable Artificial Intelligence (AI) Framework that integrates visual, historical and material data to provide accurate, transparent and reliable sculpture valuations. Sculpture Valuation is a complex and highly subjective process determined by aesthetics, historical significance, material used, condition and market trends. The current methods of evaluating sculptures; including traditional expert-based and hedonic systems; lack the consistency, transparency and ability to leverage visual and three-dimensional characteristics of sculptures. Existing systems that utilize Artificial Intelligence (AI) have focused primarily on paintings; the challenges associated with sculptural artifacts remain unresolved.
Data from various auction houses, galleries and auction records were collected in multiple ways (images, provenance, materials, past sales etc.). Visual features were extracted from these by applying computer vision analysis; historical context was established through natural language processing, physical characteristics were captured as structured data; these three sets of features were combined and processed in an ensemble based/predictive model that provides uncertainty estimates. The data used to create the ensemble model were validated by splitting them by time; standard metrics were used to assess the performance of the predictive model (accuracy and reliability). In addition, explainable models and risk modelling provide practical uses for users.
2.PROBLEM STATEMENT
5. REVIEW OF LITERATURE
The opinions of experts vary widely regarding the value of sculptures, which makes sculpture valuations very subjective and inconsistent. Most current Ai and statistical valuation methods focus solely on paintings, and these methods have a number of shortcomings; namely they do not give valuations with estimates of uncertainty or offer authenticity verification, and they are not easily accessed by people looking for professional appraisal services.
Previously, hedonic and statistical methods were the dominant art valuation techniques. These methods are based on quantitative and qualitative characteristics, i.e., attributes of an artwork (e.g., medium, size, condition) but do not account for visual and contextual factors when pricing an artwork. In recent years, researchers have employed new technologies such as machine learning, computer vision, natural language processing, and material analysis to increase the accuracy of art valuation; however these new research efforts have primarily focused on paintings. Existing research on sculpture continues to be fragmented, lacking uncertainty assessment, authenticity evaluation, and a coherent
To address these issues, there is an increased demand for a dedicated Ai-based Valuation System for Sculpture that is transparent, objective and focused on the values of sculptures.
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 898