Huan Wang1, Hong Wang1, Zhongyuan Jiang2,*, Qing Qian1, Yong Long1
CMC-Computers, Materials & Continua, Vol.80, No.3, pp. 4603-4620, 2024, DOI:10.32604/cmc.2024.053740
- 12 September 2024
Abstract Copy-Move Forgery Detection (CMFD) is a technique that is designed to identify image tampering and locate suspicious areas. However, the practicality of the CMFD is impeded by the scarcity of datasets, inadequate quality and quantity, and a narrow range of applicable tasks. These limitations significantly restrict the capacity and applicability of CMFD. To overcome the limitations of existing methods, a novel solution called IMTNet is proposed for CMFD by employing a feature decoupling approach. Firstly, this study formulates the objective task and network relationship as an optimization problem using transfer learning. Furthermore, it thoroughly discusses… More >
Login
Register
