Skip to main contentAll people

Associate ProfessorChang Xu

ARC Principal Research Fellow

Faculty of Engineering

Orcid identifier0000-0002-4756-0609 (opens in a new tab)⁠
  • ARC Principal Research Fellow
    Faculty of Engineering

Research projects & supervision summary

Project Opportunities

Title: Deep Learning for Automatic Game Character Designs

Summary of opportunity:

 

This project aims to develop personalized diffusion models that enable users to generate, edit, and interact with 3D character assets using intuitive text prompts and image uploads.

Research Area:

Deep Learning, Computer Vision

 

Opportunity synopsis:

 

The proposed research project aims to develop advanced personalized diffusion models that enable users to generate, edit, and interact with 3D character assets using intuitive text prompts and image uploads. By optimizing the diffusion process and integrating sophisticated natural language processing and computer vision techniques, the project will translate user inputs into high-quality 3D characters. An intuitive user interface will facilitate real-time adjustments, while comprehensive user studies and performance metrics will guide iterative improvements. This technology promises to enhance creative workflows in gaming, animation, and virtual reality, making 3D character generation more accessible and personalized.


This project will employ advanced mathematical modeling to optimize the diffusion process for 3D character generation. It will utilize state-of-the-art NLP techniques to interpret text prompts and cutting-edge computer vision methods to analyze image uploads. The development of an intuitive user interface will enable real-time interaction and refinement. Comprehensive user studies and defined performance metrics will be conducted to evaluate effectiveness and guide improvements.

 

Offering:
A PhD scholarship for 3.5 years at the RTP stipend rate (currently $41,753 in 2025). International applicants will have their tuition fees covered. 

 

Successful candidates must have:

  • Proficiency in programming languages like Python and experience with deep learning frameworks (e.g., TensorFlow, PyTorch) and NLP tools.
  • Prior research experience in machine learning, computer vision, and NLP is highly desirable.
  • Strong analytical skills and excellent communication abilities.
  • A passion for AI-driven technologies in creative industries is essential.

Additionally, candidates are expected to have a publication record in relevant venues.

 

How to apply: To apply, email c.xu@sydney.edu.au the following:

  • CV
  • Transcripts

RESEARCH PROJECTS & ACTIVITIES

  • RESEARCH-BASED DEGREE SUPERVISION
    A Comprehensive Exploration of Video Understanding: Perspectives on Sampling, Backbone, Representation, and Cross-Modal Learning
  • RESEARCH-BASED DEGREE SUPERVISION
    A Unified Defence Approach to the Text-to-Image model
  • RESEARCH-BASED DEGREE SUPERVISION
    Adapting Neural Architecture Search for Efficient Deep Learning Models
  • RESEARCH-BASED DEGREE SUPERVISION
    Advancing Contrastive Learning for Visual Representation: From Unsupervised to Supervised Paradigms
  • RESEARCH-BASED DEGREE SUPERVISION
    Adversarial Data Generation for Robust Deep Learning
  • RESEARCH-BASED DEGREE SUPERVISION
    Anticipating Hazards in Machine Translations of Public Health Resources via Advanced Text Classification Pipelines
  • RESEARCH-BASED DEGREE SUPERVISION
    Attention Calibration for Reducing Hallucination in Large Vision-Language Models
  • RESEARCH-BASED DEGREE SUPERVISION
    Automated Synthesis of Multi-Embodiment Dexterous-Hand Manipulation Data from Human Interaction Videos
  • RESEARCH-BASED DEGREE SUPERVISION
    Automated Synthesis of Multi-Embodiment Dexterous-Hand Manipulation Data from Human Interaction Videos
  • RESEARCH-BASED DEGREE SUPERVISION
    Boosting Adversarial Robustness via Neural Architecture Search and Design
  • RESEARCH-BASED DEGREE SUPERVISION
    Conditional Consistency Regularization for Semi-supervised Multi-label Classification
  • RESEARCH-BASED DEGREE SUPERVISION
    Controlling Generative Trajectories for Image Synthesis and Editing
  • RESEARCH-BASED DEGREE SUPERVISION
    Data-efficient Visual Recognition and Localization
  • RESEARCH-BASED DEGREE SUPERVISION
    DeCoP: Dependency Controlled Pre-training for Time Series Representation Learning
  • RESEARCH-BASED DEGREE SUPERVISION
    Deep Learning Methods for Biomedical Representation Learning and Generative Modelling
  • RESEARCH-BASED DEGREE SUPERVISION
    DeepFake's Adversary: Disrupting DeepFake by Perturbations
  • RESEARCH-BASED DEGREE SUPERVISION
    Developing an AI-Driven ESG Investment Framework: Enhancing ESG Score Prediction, Portfolio Optimization, and Financial Impact Assessment
  • RESEARCH-BASED DEGREE SUPERVISION
    Differentiable Grid Sample Pruning for Generalizable Vision-Language-Action Model
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Compression Algorithm-System Co-Design for Large-Scale Model Training and Inference
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Conditional Generation on Scale-based Visual Autoregressive Models
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Deep Neural Architecture Design and Training
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Exact and Approximate Methods for Computing Graph Edit Distance
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Neural Architecture Search with an Active Performance Predictor
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Prompt Engineering for Large Foundation Models
  • RESEARCH-BASED DEGREE SUPERVISION
    Efficient Training Data Attribution on Diffusion Models