A Comprehensive Privacy Protection Solution: Enhancing Face Obscuring Techniques with Automated Identification, Mobile Integration, and Scalable Services

Ilhwan Kim, Hyeongseok Oh, Sangwoo Yun, Heunseung Lim, Jaeseok Ryu, Joonki Paik

Abstract: We propose a novel application aimed at protecting personal privacy. With the proliferation of various media platforms, issues related to personal portraits have emerged. To address these problems, previous applications propose methods for either Mosaic or replace it with another object. However, existing methods suffer from issues such as high time costs and unnatural rendering. To address these problems, we propose designing an automatic face identification and replacement system by combining face detection, super-resolution, and face replacement models. This approach aims to reduce processing time and enhance natural rendering. In addition to the technological advancements, we suggest expanding this system into a mobile service, allowing users to easily protect their privacy on the go. This mobile platform would provide a user-friendly interface and seamless integration with various social media and communication apps, making it accessible to a broader audience. Furthermore, we explore potential expansion opportunities, such as cloud-based processing for scalability and the development of APIs that other developers can integrate into their own applications, thereby broadening the service's reach and impact. We demonstrate that our approach is simpler and more natural compared to traditional handcrafted tasks by applying it to diverse scenarios.

External IDs:dblp:conf/iccel/KimOYLRP25