3rd International Informatics and Computer Science Conference
Machine Learning for Cartoon Character Recognition
Yayıncı:
Gazi Üniversitesi
This study focuses on Convolutional Neural Networks (CNNs), a sig nificant technology that enables artificial intelligence systems to operate inde pendently or with minimal human involvement. CNNs are powerful tools used for handwriting, digit, character, and object recognition. These networks can learn and adapt to solve various problems with minimal input data. The study aims to address the challenges of using automated methods to identify and rec ognize cartoon characters. Advances in machine learning have enabled the de velopment of accurate and efficient methods for the detection of cartoon charac ters, especially using Convolutional Neural Networks (CNNs). In this study, a special CNN-based network is trained, an end-to-end architecture is designed us ing machine learning techniques, and a new algorithm is developed for the multi class classification problem. A new dataset of cartoon characters has been cre ated. Studies have shown that the caricature character recognition algorithm has a high success rate.
Yayınevi Bilgileri
Alıntı Yap
Benzer Yayınlar
Kitap Bölümleri
Bu Sayfayı Yazdırın
Endnote
Mendeley