From Image to Prompt

Yaku83

Newbie
Joined
Apr 29, 2017
Messages
5
Reaction score
0
Hi everyone, I need help understanding how to create prompts to generate images similar to the ones attached. Can anyone help me?"


Immagine 2024-12-08 170255.png
Contabile.png
 
Most AI image platform offers a public gallery where people can not only see the output but the details about the image, including the prompt used to generate it.

Additionally, I search on Google for "From Image to Prompt" and, to my surprised, there are multiple free online tools where you can upload the image and it will generate a suggested prompt you can use to potentially replicate that image.

After a few testing, this is the tool I believe generated the best output.

https://imageprompt.org/image-to-prompt

By the way, I use this image for testing:

113e9246f8c9565e98553763b4591f8f.png


And this is the prompt it generated:

A middle-aged light-skinned man, centered in the image, is depicted in a cartoonish, stylized manner. He wears a light-colored, button-down collared shirt and a dark tie. His facial expression is one of intense worry or concern, with wide eyes and an open mouth. He is wearing round-framed glasses. His body posture is tense, with fists clenched, suggesting a state of agitation or readiness. The image's setting appears to be an office or similar professional environment. The background is a subdued, neutral tone, with a suggestion of a chair or desk visible. Large, bold yellow text, "CONTABILE," is prominently displayed over his upper body. The overall tone is somewhat dramatic and animated. The style is stylized 3D cartoon. The lighting is bright, with highlights on the face and clothing details, and the colors are somewhat muted, but clear. The perspective is slightly from above, looking down at the subject. The composition emphasizes the subject's posture and expression, with the bold text drawing the viewer's eyes to the subject's reaction. Small details are visible, such as buttons on the shirt and the frame of the glasses.

So, I got curious and tested it that prompt on 3 free AI image generation apps: Leonardo, Ideogram, and PicLumen, and here are the results:

For Leonardo (Seed #: 17491045)...
4ed66ead3f548f430169aa1353869228.jpg
fe74f1e861dafebad49fabd0572228e6.jpg
d00e1abe574415da0e567080a8a2729f.jpg
926fe579e6c69466a0a83a1c8f3b512a.jpg


For Ideogram (Seed #: 220286080)
58f32037a973153d6e255b797eff8567.webp
84b8fb541fc9f0c1715389ae5c5bd4b4.jpg

6679f8703dff4289b74a1df078a43070.jpg
d39291d7252aee170689b1f34fdaa915.jpg


For PicLumen...
f61e8b9320c777f887365d459fde7f1f.png


I hope this help, cheers!
 
You should avoid this style, it's universally hated by everyone on the Internet
 
Interesting. I tried the tool from Roger with the same pic for flux model. And generated images with the image generator inside. Here is what i got:
113e9246f8c9565e98553763b4591f8f.png
ai_g_img_2024-12-12_1734011626528_8297.jpg
generated by flux 1.1 pro
Prompt: Middle-aged man, light skin tone, wearing a light gray/white button-up shirt and dark-colored tie, with a serious and worried expression. He's clutching both fists tightly, and his eyes are wide with concern. Round-framed glasses are perched on his nose. The subject's pose and expression depict a sense of anxiety or apprehension, a perspective of close-up on his upper body. The setting appears to be an office environment. An armchair, light beige or tan color, is partially visible in the background. Large, bold, bright-yellow text, "CONTABILE," is prominently displayed over the image. The text is in a sans-serif, bold font. The lighting is seemingly natural or studio lighting. Camera angle is a direct, centered shot. Realistic lighting on the face is clearly visible, a realistic style is conveyed in the image. Cartoonish but somewhat realistic style. Animated, yet photorealistic. Close-up view. Calm yet tense visual.

immagine-2024-12-08-170255-png.402487
ai_g_img_2024-12-12_1734012218439_940.jpg
flux 1.1 pro
A powerful illustration of an elderly man cradling a young child. The man, middle-aged, with graying, slightly long hair, and visible wrinkles, expresses profound sorrow and weariness. His gaze is downcast, reflecting a heavy emotional burden. He has a weathered, somewhat gaunt appearance. His expression is somber, reflecting profound sadness and deep empathy. He is dressed in a simple, olive-green or khaki-colored, long-sleeved button-down shirt, with visible texture suggestive of a rugged fabric. The child, with a similar somber facial expression, appears to be a pre-teen, likely around 8-10 years old. The child's attire is a simple dark, likely blue shirt and shorts. Their eyes are reflecting the pain felt by the elder. The composition is dramatic, with the old man's shoulders filling most of the frame. The perspective is close-up, emphasizing the emotional connection between the figures. Muted, earthy tones dominate the environment, with a background hint of a fire or stormy scene. The background subtly suggests a desolate or dangerous environment, evoking a mood of hardship and survival. A title, "CAMMINARE," rendered in a bold, capital typeface, is prominent. The lighting is dramatic, highlighting the figures and emphasizing the somber mood. The overall style is realistic, aiming for a heightened sense of emotion, rather than a specific artistic movement. The illustration uses careful attention to detail, especially to facial expressions and texture. Deep shadows deepen the somber atmosphere, focusing attention on the heavy emotional burden of the subject.

Funny.
 
Give the image to ChatGPT and ask it to extract the prompt used for the image.
 
Back
Top