A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Gemstone Classification Using Deep Convolutional Neural Network
Gemstones are precious minerals or rocks that can be cut, faceted, and polished to be used in jewelery. Gemstones are valuable commodities in today’s market. For thousands of years, gemstones have been prized for their beauty, metaphysical properties, and commercial applications. The task of classifying a gemstone is very complicated because of its nature, which is associated with various aspects. This paper proposes a convolutional neural network (CNN) based model to automatically classify gemstones. In addition, the data augmentation (DA) technique is employed to expand the quantity of data by incorporating slightly altered replicas of available data. Image cropping has been introduced to remove unwanted elements from the image. For the experiment, we have used the DenseNet169 pre-trained transfer learning (TL) model, and the experiment was conducted on a dataset that contains 87 different categories of gemstones. Our model has been evaluated against all existing models on the following metrics: accuracy, loss function, recall, precision, and F1 score. The proposed model has 79% accuracy (ACC), which shows significant improvement over existing models.
Gemstone Classification Using Deep Convolutional Neural Network
Gemstones are precious minerals or rocks that can be cut, faceted, and polished to be used in jewelery. Gemstones are valuable commodities in today’s market. For thousands of years, gemstones have been prized for their beauty, metaphysical properties, and commercial applications. The task of classifying a gemstone is very complicated because of its nature, which is associated with various aspects. This paper proposes a convolutional neural network (CNN) based model to automatically classify gemstones. In addition, the data augmentation (DA) technique is employed to expand the quantity of data by incorporating slightly altered replicas of available data. Image cropping has been introduced to remove unwanted elements from the image. For the experiment, we have used the DenseNet169 pre-trained transfer learning (TL) model, and the experiment was conducted on a dataset that contains 87 different categories of gemstones. Our model has been evaluated against all existing models on the following metrics: accuracy, loss function, recall, precision, and F1 score. The proposed model has 79% accuracy (ACC), which shows significant improvement over existing models.
Gemstone Classification Using Deep Convolutional Neural Network
J. Inst. Eng. India Ser. B
Chakraborty, Bidesh (author) / Mukherjee, Rajesh (author) / Das, Sayan (author)
Journal of The Institution of Engineers (India): Series B ; 105 ; 773-785
2024-08-01
13 pages
Article (Journal)
Electronic Resource
English
Gemstone Classification Using Deep Convolutional Neural Network
| Springer Verlag | 2024
Mineralizing agent for regenerated gemstone, gemstone regeneration method and regenerated gemstone
| European Patent Office | 2023
Automatic classification of pavement crack using deep convolutional neural network
| Taylor & Francis Verlag | 2020
Reinventing the Gemstone Market
Online Contents | 2017