VGU RESEARCH REPOSITORY
Please use this identifier to cite or link to this item:
https://epub.vgu.edu.vn/handle/dlibvgu/1757
Title: | A study on AI technologies and application to human detection in animated movies | Authors: | Nguyen Tuan Ngoc | Keywords: | Supervised learning;Duplicate image filtering;Transfer learning | Issue Date: | 2023 | Abstract: | In the realm of human detected in animation, the challenge often revolves around large data availability. This thesis, titled "A Study on AI Technologies and Application to Human Detection in Animated Movies" delves into this issue by harnessing the power of transfer learning. A key innovation is the introduction of the duplicate image filtering technique, designed to enhance data quality for large-scale datasets by identifying and moving similar images within clusters. This function is expedited using advanced concurrency mechanisms, thereby accelerating the data preparation phase and ensuring robustness in human detection models. Moreover, the study proposes an inventive approach that combines supervised learning with image super-resolution, drawing inspiration from Real ESRGAN. This amalgamation serves a dual purpose: enhancing the accuracy of human detection and seamlessly integrating these detected figures into various animated contexts |
URI(1): | https://epub.vgu.edu.vn/handle/dlibvgu/1757 |
Appears in Collections: | Computer Science (CS) Computer Science (CS) |
Files in This Item:
File | Description | Size | Format | Existing users please Login |
---|---|---|---|---|
A study on AI technologies and application to human detection in animated movies.pdf | 6.94 MB | Adobe PDF |
Items are protected by © Copyright of Vietnamese - German University Library