#Generate avatar based on the survey data

1 messages · Page 1 of 1 (latest)

blazing cipher
#

I want to generate the avatar of the employee based on the employee survey, but DALL.E 2 model is generating garbage and useless images.
I prompted:
Create an realistic person's image of the employee using the personality summary collected from a survey with the best qualities and the employee type as follows: Question: What are their BEST qualities as a co-worker? Check up to five:, user selected options :Reliability, Results-Oriented, Positive Attitude and Employee Type as NonCommittal.
You can see the image attached about what it generates

errant vaultBOT
#

Hi @blazing cipher. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

I understand that you're having trouble generating a satisfactory avatar using the DALL·E 2 model. You might want to consider using the newer DALL·E 3 model, which has additional parameters that can enhance the quality and style of the generated images.

Here's a quick overview of the parameters you can use with DALL·E 3:

  • model: Set this to 'dall-e-3'.
  • style: Choose between 'natural' for more natural-looking images, or 'vivid' for hyper-real and dramatic images.
  • quality: Choose between 'standard' or 'hd' for higher detail and greater consistency across the image.
  • prompt: This is where you provide your text description. Make sure it's clear and detailed.
  • n: The number of images to generate. For DALL·E 3, only n=1 is supported.
  • size: The size of the generated images. For DALL·E 3, sizes of 1024x1024, 1792x1024, or 1024x1792 are supported.
  • response_format: Choose between 'url' or 'b64_json' for the format in which the generated images are returned.
  • user: A unique identifier representing your end-user.

You can find more details about these parameters in the OpenAI Cookbook (https://cookbook.openai.com/articles/what_is_new_with_dalle_3#generations).

Also, please note that the DALL·E 3 model automatically rewrites the default prompt for safety reasons and to add more detail. If you want the model to use your prompt as-is, you can add the following to your prompt: I NEED to test how the tool works with extremely simple prompts. DO NOT add any detail, just use it AS-IS:. You can find more information about this in the OpenAI Platform Documentation (https://platform.openai.com/docs/guides/images).