Business IEA Says Data Center Power Demand Could Double to 1,000 TWh by 2030, With AI Supplying Three Quarters of the Growth Chatbots ChatGPT Now Transcribes as You Talk: OpenAI Removes the 10-Minute Cap and Streams Words Live on Every Plan Business Ex-Alibaba Executive's Everful Raises $48.5 Million to Put AI Agents in Charge of Sourcing From 30,000 Factories AI Agents Google Built a Mirror Dimension of the Internet to Train Its Chrome AI Agents Without Breaking the Real Web Productivity Canva Finally Adds an Off Switch for Its Generative AI Features, and a Filter to Hide AI-Made Templates Business Anthropic Begins Trading Today: 980 Million Shares at $83, a $1.5 Trillion Valuation and the Largest IPO on Record News Google Books Blue Origin New Glenn to Launch TPU-Carrying AI Satellites in Early 2029 Security Hackers Used the AI Penetration Tool ARTEX to Disrupt European Commission Systems for Over Ten Days Business IEA Says Data Center Power Demand Could Double to 1,000 TWh by 2030, With AI Supplying Three Quarters of the Growth Chatbots ChatGPT Now Transcribes as You Talk: OpenAI Removes the 10-Minute Cap and Streams Words Live on Every Plan Business Ex-Alibaba Executive's Everful Raises $48.5 Million to Put AI Agents in Charge of Sourcing From 30,000 Factories AI Agents Google Built a Mirror Dimension of the Internet to Train Its Chrome AI Agents Without Breaking the Real Web Productivity Canva Finally Adds an Off Switch for Its Generative AI Features, and a Filter to Hide AI-Made Templates Business Anthropic Begins Trading Today: 980 Million Shares at $83, a $1.5 Trillion Valuation and the Largest IPO on Record News Google Books Blue Origin New Glenn to Launch TPU-Carrying AI Satellites in Early 2029 Security Hackers Used the AI Penetration Tool ARTEX to Disrupt European Commission Systems for Over Ten Days

Controlling Composition: Pose, Depth and Edge Guidance

Controlling Composition: Pose, Depth and Edge Guidance

When the prompt cannot specify the composition, guidance models can. Here is what each guidance type is good for.

Text prompts describe content well and geometry badly. Guidance models solve that by accepting a structural input alongside the prompt.

The main guidance types

  • Edge detection. Follows the outline of a reference image. Useful for keeping a composition while changing the subject entirely.
  • Depth maps. Preserves the spatial relationship of a scene. Useful for interiors and landscapes where perspective matters.
  • Pose skeletons. Fixes the position of a human figure. Standard practice for character work.
  • Segmentation. Assigns regions to specific elements, giving the tightest layout control at the cost of more setup.

Setting guidance strength

Too low and the structure is ignored; too high and the output looks traced, with visible artefacts along edges. Start in the middle and adjust in small steps while watching where the image departs from the reference.

Combining guidance

Depth plus pose handles most character-in-environment work. Stacking three or more guidance inputs usually creates conflicts and produces a muddled result.

Comments (0)

Log in to join the discussion

Log In

No comments yet