A5003087973- Handwritten Text Recognition Techniques
- Generative Adversarial Networks and Image Synthesis
- Human Motion and Animation
- Aesthetic Perception and Analysis
- Digital Media and Visual Art
- Color perception and design
- Image Processing and 3D Reconstruction
- Topic Modeling
- Natural Language Processing Techniques
Xi'an Jiaotong University
2022-2024
Even though graphic layout generation has attracted growing attention recently, it is still challenging to synthesis realistic and diverse layouts, due the complicated element relationships varied arrangements. In this work, we seek improve performance of by incorporating concept regions, which consist a smaller number elements appears like simple layout, into process. Specifically, leverage Variational Autoencoder (VAE) as overall architecture decompose decoding process two stages. The...
During the creation of graphic designs, individuals inevitably spend a lot time and effort on adjusting visual attributes (e.g., positions, colors, fonts) elements to make them more aesthetically pleasing. It is trial-and-error process, requires repetitive edits, relies good design knowledge. In this work, we seek alleviate such difficulty by automatically suggesting aesthetic improvements, i.e., taking an existing as input generating refined version with improved quality output. This goal...