While written prompts express creative ideas, specific placement of the visuals that is accurate and precise can still be challenging. An object may be too far away, too close to another object, or out of the composition. Creating or editing images with Dreamina AI gets extra visual cues from simple doodles and shapes.

Sketching and describing things at the same time provide more detail on object placement, proportions, and relationships to other elements around the object. Circles can indicate areas to be addressed, arrows can convey direction, and approximate outlines can define boundaries. These markings direct creative intent, but do not necessarily ensure exact reproduction. Dreamina AI combines image creation and editing capabilities, enabling rough visual instructions to be polished up into sophisticated compositions.

Why Visual Guidance Can Improve AI Image Instructions

Concepts, materials, colors, and artistic styles are communicated by text prompts, and sketches depict approximate spatial relationships. Words like above, behind, or nearby can have multiple meanings. It is helpful to show an object’s position in relation to other subjects in a rough drawing. This means that visual guidance is helpful for posters, product displays, illustrations, and background edits. Dreamina AI’s AI Image generator workflow involves using descriptive prompts and other available tools, thereby generating results that might not be exactly what was initially sketched.

Choose the Right Marking Method for the Intended Change

Various markings convey a variety of visual intentions. Doodles represent a general idea of the shape of an object, and arrows show direction, motion, or connections between objects. The placement areas are indicated by bounding boxes, while circles indicate areas that need attention. Points specify a location; coordinates provide location information only if the workflow selected supports them. These markings are intended as visual marks only and are not a technical mask. It depends on the tool that they use, the prompt that was given to them, and the editing context.

Steps: Guide Results With Doodles and Shapes in an AI Image Generator

Step 1: Prepare your doodle and write a matching prompt

  1. First, sign up or log in to Dreamina, and keep your doodle or shape sketch ready.
  2. Next, go to the “Create” tab in the left menu and select the “AI Image” option.
  3. Later, write a detailed prompt that explains your doodle, along with image elements, theme, and concept.

Step 2: Add your sketch, choose a model, and generate

  1. Next, pick an image model from the options at the bottom of the dialogue box.
  2. Choose Seedream, GPT Image, or Nano Banana, based on how closely you want it to follow.
  3. Upload your doodle as a reference image so the result looks more precise and natural.
  4. Then, match your sketch layout by setting the “Aspect ratio” and “Resolution” tabs.
  5. Pick the number of outputs in “Images to generate”, and set a custom size with “Dimensions”.
  6. Place a title or label on your artwork using the “Draw text on image” tab.
  7. Finally, click the “Generate” tab to turn your shapes into finished visuals.

Step 3: Polish the result and export it

  1. You will see several drafts, and you can use “Edit prompt” or “Regenerate” whenever the shapes drift from your sketch.
  1. Enhance clarity in your favorite draft through the “Creative upscale” option.
  2. Use “Retouch” to correct shape edges, clean up messy areas, and enhance fine details.
  3. Take out unwanted marks with “Remove”, or enlarge the scene using the “Expand” tab.
  4. You can easily convert the image you generated into a video through the “Generate video” tab.
  1. Dreamina gives you complete freedom to shape the image however you like.
  2. Finally, click the “Download” tab at the top right to save the image on your device.

Combine a Sketch With a Clear Written Prompt

Visual marking is most effective when it is used in conjunction with, but not in lieu of, written instructions. Draw a sketch to show where to place the object and describe its appearance, size, color, fabric, and style. Describe how it should relate to other subjects and backgrounds. For instance, draw a lamp next to a chair, then indicate that it is made of metal and that it emits a warm light. Or, draw a blank product display case and tell what to put in. Drawings should be simple and original elements should be specified as to what should be kept consistent.

Six Practical Ways to Direct AI Images With Visual Marks

  • Object placement: Identify where the object is to be placed, then describe how the object looks and how it relates to other objects around it.
  • Composition planning: Draw out general locations of subjects and explain foreground and background, scale, and hierarchy.
  • Describe the background changes: Mark the target area and describe the scenery, lighting, and surroundings.
  • Design corrections: Circle the element that has been designed in the wrong way and describe how it should be designed, what color, what shape, etc.
  • Poster layout: Mark off areas where headlines, illustrations, and supporting graphics will be located and specify text and hierarchy.
  • Illustration development: Draw down key silhouettes and explain the final illustration style, textures, and visual treatment.

Use Spatial Guidance for More Organized Design Compositions

Spatial guidance is used to organize subjects, graphics, and negative space in a composition. Clearly defined placement in a product display, readable text, and a balanced visual hierarchy in a poster. These relationships can be conveyed through sketches before going into the detailed rendering process. But there should be no conflicting directions between written instructions and markings. Check the text created for spelling, alignment, and readability. Although visual guidance helps to communicate intent, there is still a need for the final layout review prior to publication.

Correct Misinterpreted Sketches and Unwanted Changes

Ambiguous marks may yield surprising forms or perplexing placements. Make the sketch simpler and more precisely specify the desired element in writing. If placement appears to be wrong, modify the outline and indicate its location next to other subjects. Support original aesthetics if changes of style to original elements are undesirable. When the composition is busy, eliminate any unneeded objects. If the existing details are distorted, go back to the reference image and use supported editing controls. If the overall interpretation is not suitable, then re-generate and carefully check the revised output.

From Rough Visual Directions to a Finished Image

First sketches can be developed into successive generations and prompt revisions. Every iteration can help to define placement, enhance proportions, and adjust the foreground/background relationship. If only a detail will need to be edited out, do so without having to build a satisfactory composition. Use “Creative upscale” where the higher resolution will enhance the presentation. Check edges, text, spacing, and consistency throughout the image before exporting. The review of the result can help identify issues which may not have been anticipated by the original instructions.

Conclusion

Another means of conveying spatial concepts, in addition to written prompts, is doodles and shapes. Dreamina AI offers a creative workflow with image references, descriptive instructions, and available editing controls. Simple drawings, exact descriptions, and careful reviews clarify the desired outcome. The best visual guide is the one that is clear and unambiguous and is able to convey exactly what needs to change.