Custom Dataset Training with StyleGAN

Job ID: 38772633

Budget: $30 – $250 USD

I have a custom dataset of face images that I would like to train using StyleGAN to generate fake images for research purposes.

Key Requirements:
1. Implement both manual and automated control of noise, allowing me to see the differences in the output
2. Utilize two types of discriminators, comparing their results and noting any discrepancies
3. Apply CLIP and another similar tool for analysis and comparison
4. Use CMA optimizer and another similar tool for further comparison and evaluation

Ideal Skills:
- Proficiency in using StyleGAN
- Experience with image generation and manipulation
- Familiarity with various types of discriminators
- Knowledge of CLIP and CMA optimizer
- Ability to implement both manual and automated noise control

I am looking for someone who can meet these requirements and has the necessary skills and experience to carry out this project successfully.
Related categories: Python Machine Learning (ML) Pytorch StyleGAN