Hey everyone! 👋
I’m planning to create a practical tutorial series focused on dataset preparation, especially for training generative AI models. The goal is to walk you through my entire workflow from extracting images directly from videos to processing and optimizing them for consistent results.
📌 Topics I’ll Cover:
✔ Frame extraction from anime/videos etc.
✔ Character Extraction & Classification
✔ Semi-automated processing: Tools and scripts for cleaning, cropping, and adjusting images.
✔ Dataset filtering and organization to improve training outcomes.
✔ Tips for maintaining quality and consistency (resolution, lighting, style, etc.).
To make it more practical, I’ll use an anime or manhwa as an example. I'll use the anime you suggest in this post as my reference material
The final dataset will be released alongside the trained models for public use.
📢 What Do You Think?
🔹 Which topic interests you most? Should I dive deeper into any specific step?
🔹 Would you prefer these tutorials in English or Spanish?
Drop your thoughts and suggestions below!