A5013302961- Generative Adversarial Networks and Image Synthesis
- Digital Media Forensic Detection
- Advanced Image and Video Retrieval Techniques
- Multimodal Machine Learning Applications
- Advanced Image Processing Techniques
- Video Analysis and Summarization
Guangdong University of Technology
2021-2023
Guizhou University
2023
Text-guided image inpainting aims to complete the corrupted patches coherent with both visual and textual context. On one hand, existing works focus on surrounding pixels of without considering objects in image, resulting characteristics described text being painted non-object regions. other redundant information may distract generation interest restored image. In this paper, we propose an adversarial learning framework mask reconstruction (ALMR) for guidance, which consists a two-stage...
In this paper, we propose a multi-branch semantic learning network (MSLN) to generate image according textual description by taking into account global and local semantics, which consists of two stages. The first stage generates coarse-grained based on the sentence features. second stage, fine-grained generation model is constructed inject sentence-level word-level semantics images attention modules, textures, respectively. particular, devise channel fusion module (CFM) fuse features in...