Machine Vision for Analyzing Contemporary Sculptures
Bibliographic Data
| ID | 22196472 |
|---|---|
| Authors | Salai Tamilarasan S (SRM Institute of Science and Technology), Thakur Gurjeet Singh (0000-0003-2979-1590, Chitkara University), Takveer Singh, Mukul Pandey (0000-0002-3384-2701, National Institute of Technology Jamshedpur), Sonia Pandey (0000-0001-6661-5823, Noida International University), Kalpana Munjal (Vivekananda Global University), Gopinath K Gopinath K, Gopinath Achary K (Vinayaka Missions University), Prince Kumar (0000-0001-8933-7571, Noida International University), Dipti Nitin Dixit (International Institute of Information Technology) |
| Year | 2025 |
| Volume | 6 |
| Issue | 4s |
| Pages | 484-494 |
| Publication date | 2025-12-25 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | ShodhKosh: Journal of Visual and Performing Arts (JOURNAL) |
| Journal identifiers | ISSN: 2582-7472 • E-ISSN: 2582-7472 |
| Publisher | Granthaalayah Publications and Printers (PUBLISHER • IN) |
| DOI | 10.29121/shodhkosh.v6.i4s.2025.6849 |
| OpenAlex | W7117464223 |
| Language | EN |
| References cited | 14 |
Incorporation of machine vision in the current analysis of sculpture brings a paradigm shift in the art documentation, interpretation, and preservation. This paper presents computational paradigms that incorporate state-of-the-art techniques of computer vision with aesthetic criticalism to assess sculptural structure, materiality and style. The study uses very high resolution images and 3D scans of contemporary sculptures to develop an elaborate pipeline that includes segmentation, normalization and annotation based on metadata. Deep learning models used in methodology include Convolutional Neural Networks (CNNs), 3D-CNNs, and PointNet which are used to extract multidimensional features including surface curvature, geometric and textural complexity. These properties make it easy to classify objects objectively and interpret them subjectively according to the rules of art. Accuracy, mean intersection over union (mIoU), and feature consistency metrics are some of the evaluation metrics that measure the accuracy of the model prediction. Findings reveal an effectiveness of machine vision to identify fin-grained sculptural dimensions, help improve curation, automate restoration, and integrate museums virtually. In addition, the research highlights the interdisciplinary prospects of using art history, computational design, and artificial intelligence to learn more about the meaning and development of art. This study will help fill the gap between visual computing and creative analysis by offering a repeatable structure to digitize and interpret three-dimensional art, which will open the way to new directions of digital heritage, teaching visualization, and aesthetic value analyses with the help of AI
Convolutional neural network · Deep learning · Machine Vision · Sculpture · Visualization · Aesthetic Perception and Analysis · Art History and Market Analysis · Digital Media and Visual Art
| Citation velocity | historical |
|---|---|
| Highly cited | No |