Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

Video course · 14 chapters · 31 min · certificate

Edit Photos by Instruction in ComfyUI with Qwen Image

SECourses teaches how to edit photos with natural-language instructions inside ComfyUI using the Qwen Image model. The chapters cover denoise comparison, two-reference edits, character consistency, storyboards, and relighting.

Become a memberAlready a member? Sign in

What you'll learn

  • Compare image-to-image and conditioning-based editing approaches in ComfyUI
  • Measure changed pixels using difference masks and threshold nodes
  • Replace text on photographed signs with exact characters and punctuation
  • Transfer clothing between two reference images using named tags
  • Build a multi-scene storyboard with one consistent character
  • Apply relighting instructions that follow photographic light descriptions

Chapters

14 chapters · 30:48
  1. 1:48 01ComfyUI Update Members Edit photos by instruction: today's results Update ComfyUI with the same route from Lecture 2 before starting the lesson.
  2. 3:02 02Load Image Members Turn text-to-image into image-to-image Replace the Empty Latent Image with a Load Image node and encode the photo into a latent.
  3. 3:04 03Denoise Grid Members Compare four denoise strengths in one grid Run four samplers with different denoise values and collect results in one image grid.
  4. 2:06 04Conditioning Edit Members Edit with the photo in the conditioning Use the photo as a reference inside the text encoder for instruction-based editing.
  5. 1:14 05Regeneration Members Why denoise one is still editing Conditioned regeneration redraws the whole picture, so unchanged areas are near-identical, not identical.
  6. 4:02 06Difference Map Members Measure changed pixels with a difference map Subtract the original from the edit, threshold the result, and save a white map of changed pixels.
  7. 2:14 07Resolution Members Pixel budget: detail versus time Compare 1024 and full-resolution reference encoding to see the detail-time trade-off.
  8. 1:29 08Text Editing Members Replace words on a photographed sign Edit sign text in a bakery storefront photo with exact quotes and characters.
  9. 1:30 09Two References Members Two references with named roles Use two Load Image nodes with named tags to transfer a jacket from one photo to the man in another.
  10. 1:22 10Turnaround Members One character: a turnaround sheet Generate a three-view turnaround sheet of the explorer from one anchor photo.
  11. 3:24 11Storyboard Members Place the explorer in three storyboard scenes Place the explorer into railway, hilltop, and workshop scenes with the anchor as a named reference.
  12. 3:22 12Sequential Edits Members Five sequential edits and a clean rollback Run five edits one after another on the latest result, then recover by returning to the clean anchor.
  13. 1:05 13Relighting Members Relight a portrait by instruction Describe light the way a photographer would to relight the portrait.
  14. 1:06 14Export Members Export the workflow and recap Export the workflow file and review the two photo-editing methods learned.

Jobs this course suits

Our AI checked this course against 500 jobs; these get the most out of it. Each job links to its learning path.

PhotographersGraphic Designers