Workflow Goal
Provide a clean reference implementation for Anima-based image editing.
Use this workflow as a starting point and adapt the edit instructions to match your preferred artistic direction.
Workflow Type
Image-to-Image (Instruction-based Editing)
This workflow demonstrates instruction-based image editing using the YAAE (Yet Another Anima Edit) LoRA.
YAAE v0.1 is described by its author as a proof-of-concept for the current training pipeline. As with many early image editing models, results may vary depending on the input image, edit instruction, and random seed.
Observed Behavior
✓ Character identity, facial features, and overall pose are generally preserved very well.
✓ Localized edits (such as clothing color, object replacement, and background modification) produce good results in most cases.
✓ Short, natural-language edit instructions generally provide the most consistent results.
As this is a proof-of-concept image editing model, results may vary depending on the input image, edit instruction, and random seed.
Certain edits may also influence nearby semantically related regions. For example, changing a clothing color may occasionally affect nearby accessories or hair color. Likewise, large scene edits (such as replacing the entire background) may preserve the original composition while producing minor perspective or object placement inconsistencies.
If the desired edit is not achieved on the first attempt, consider:
refining the edit instruction
trying a different random seed
regenerating the image.
Overall, the workflow provides a strong foundation for exploring instruction-based image editing with Anima and serves as an excellent starting point for further experimentation.
Version
Version 1.0
Created as a minimal reference workflow for getting started with image editing using the Anima model family.
Purpose
✓ Educational
✓ Minimal
✓ Easy to extend
Verified
✓ Image-to-Image Editing
Verified Settings
✓ Steps: 25
✓ CFG: 4.0
✓ Sampler: sa_solver_pece
✓ Scheduler: Simple
These settings were used to validate this workflow and are recommended as the baseline configuration.
Verified Resolution
✓ 832 × 1216 (SDXL HD Portrait)
✓ 592 × 1024
Hardware Tested
✓ NVIDIA RTX 2080 (8 GB VRAM)
✓ 64 GB System RAM
✓ ComfyUI 0.27.0
Next
□ Image-to-Video Editing
□ Image Upscaling