| Autor: |
Kumar, A., Christy, S., Logapriya, E., Roshan, S., Selvaperumal, S. K., Kumar, P. N. |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2024, Vol. 3161 Issue 1, p1-7, 7p |
| Abstrakt: |
This proposed work aims to enhance the accuracy of bird species classification using multi-layer perceptron (MLP) in comparison with novel convolutional neural networks (CNNMaterials and Procedures: For researchers and visitors alike the categorization of birds is a significant issue. This study contains two groups i.e. Multi-layer Perceptron and Neural Convolutional Network. For every group there is a sample size of 250 images from a total sample of 70626 images of 450 bird species dataset collected from Kaggle (N=10). Using ClinCalc, the research parameters are an alpha value of 0.05, a beta value of 0.2, and a G-power value of 0.8. Results: In bird species categorization, Convolution Neural Networks are 85.84% more accurate than Multi-Layer Perceptron (65.63%). The significance value for performance is 0.000 (Independent variable T-test p<0.05), indicating statistical importance. Conclusion: The Novel Convolutional Neural Network Model outperforms the MLP classifier in the examination of bird species categorization. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
