ChatGPT Image Description and Generation
€20-50 EUR
Paid on delivery
Creating Perfect Prompts for Generating Images in ChatGPT
Project Description
I have a web application where users upload a selfie and select one term from each of the two available lists of themes. Some of the available themes include: Owl eyes, Always late, Selfie guru, Retro soul, Hedonist, Excel Expert, Always working, True friend, Cute, Comedian, Karaoke champion...
Your task is to create two precise prompts for ChatGPT that will deliver satisfactory results:
Image Analysis Prompt – A detailed analysis of the user's selfie.
Image Generation Prompt – Generating a new image based on the analysis and the user’s selections.
Task Details
1. Image Analysis Prompt
The first prompt should provide a detailed description of the user's selfie photo, including:
Gender of the person in the photo (male/female).
Facial expression (emotions, specific features).
Hairstyle (style, color, length).
Clothing (colors, patterns, accessories).
Gestures (if the person is holding something, pointing, etc.).
Age (child or adult).
Any other relevant visual information that contributes to accurate image creation.
Example result format:
Gender: Male
Facial Expression: Smiling, cheerful
Hairstyle: Short, brown, slicked back
Clothing: T-shirt, blue, no accessories
Gestures: Right hand pointing at the camera
2. Image Generation Prompt
The second prompt should use the information from the first prompt to create a new image based on the following specifications:
Image Split: The image should be divided vertically into two equal halves.
Left Side of the face should reflect the theme selected from the left list ($left_choose) with a pink background.
Right Side of the face should reflect the theme selected from the right list ($right_choose) with a red background.
Consistency with the original selfie: The face must remain recognizable and consistent with the original selfie but adapted to the themes.
Detailed personalization: Pay special attention to the hairstyle, gestures, clothing, and facial expressions to ensure the generated image aligns with the original photo and the chosen themes.
Artistic Filter: Apply one artistic filter to the entire image, such as tapestry, comic, oil painting, manga, fantasy, or abstract.
Restrictions:
The image must not contain text, icons, religious/national symbols, drugs, alcohol, or cigarettes.
The face should remain that of one person (not two different people).
Each side of the face should visually reflect the chosen theme but remain cohesive as one image.
Example:
The user selects "Hedonist" (left side) and "Karaoke champion" (right side). The left side of the image has a pink background and reflects a hedonistic style, while the right side has a red background and expresses traits of a karaoke champion.
Additional Notes:
Data from the first prompt will be used in the variable $imageTextData.
The left and right themes will be dynamically inserted into the variables $left_choose (left side) and $right_choose (right side).
Why I Need Help
Currently, I’m using the following prompts, but I am not satisfied with the results:
First Prompt (Image Analysis):
"Analyze the image and provide a detailed description of the person's facial expression (emotions conveyed, specific features), body language (posture, gestures), hairstyle (length, color, style features), clothing (color, style, patterns, accessories), identify their gender (male or female), skin color and specify if they appear to be a child or an adult."
Second Prompt (Image Generation):
"You’re an imaginative digital artist with extensive experience in creating visually compelling images based on detailed descriptions. Your specialty lies in transforming concepts and emotions into captivating visuals that resonate with the viewer.
Your task is to generate an image based on the provided selfie and the specific instructions given. Here are the details to keep in mind:
$imageTextData
Image Specifications:
Vertical Division: Divide the image vertically into two equal halves, featuring only one person. The face should be split into left and right sides, each side reflecting a different theme, not two separate people.
Left Side of Face: Apply the $left_choose theme with a light pink background.
Right Side of Face: Apply the $right_choose theme with a clean red background.
Restrictions: Ensure no symbols (e.g., no smoking, no alcohol), brands, food, text, or religious items are present.
Goal: Illustrate a single face divided into two distinct themes, where each side reflects the unique emotion or persona described in $left_choose and $right_choose."
Objective of This Task
I need prompts that will generate high-quality images that meet all the above requirements.
If you have experience with crafting ChatGPT prompts, artistic image generation, and adapting to detailed requests, I look forward to your application.
Project ID: #38779212
About the project
19 freelancers are bidding on average €40 for this job
As a versatile and experienced web developer, I am excited to leverage my wide range of skills to make your ChatGPT image description and generation project a resounding success. My proficiency in blockchain developmen More
Hi there, I’ve carefully reviewed the requirements for your GenAI project and I’m confident that my expertise in building NLP pipelines using Hugging Face and LangChain can meet your expectations. My experience includ More
Hi there, I am an ML engineer. I can start right away and deliver within the deadline. So, Let’s have a quick conversation. I can be more specific once we get all the requirements and information required to execute th More
Hi, Thank you for the clear project details about the generation of the images. I can help with the best prompt generator for creating such images. I specialize in Prompt Engineering and I do this project correctly fo More
I am the best for the job, I think and believe that even though I don't have a team, much less have the experience, I know that at the first opportunity I will leave you speechless, just trust me.
As a seasoned SEO and SMM expert, I've honed my skills in transforming concepts into visually compelling content that resonates with viewers - a crucial skill I believe to be essential for the task at hand. I am adept More
Hello, I am Khuzaima Altaf Zahid, a data scientist with two years of experience specializing in machine learning, deep learning, and web scraping. My expertise includes: Machine Learning: Proficient in classification More
Hi there, To enhance your web application for generating images from selfies, I propose refining the prompts used for image analysis and generation. The new prompts will provide detailed descriptions of the user's sel More
As an experienced AI Prompt Engineer, I'm applying to help improve the image generation process for your web application. The goal is to create two precise prompts for ChatGPT: one for analyzing a user-uploaded selfie More
As a Prompt Engineer, I leverage my expertise in crafting effective prompts to train and refine action models. With a solid background in large language models (LLMs), I focus on writing, implementing, and testing thes More
Dear [Client's Name], We are excited to offer our professional digital art creation services tailored to meet your specific needs. We specialize in crafting visually compelling and meaningful artwork that aligns perfe More
