AI-Assisted Composite Photo Generation project
Budget: $30 – $250 USD
https://drive.google.com/drive/folders/1z87tp-MYv4HxqAzJkcbQcLlf4PZGwupX?usp=sharing
AI Multi-Angle Human Model Generation Requirements
Final Deliverable:
Create four AI-generated composite photos of a specified model wearing a specified product, showing the following angles:
• Front view
• 45-degree angle view
• Side view
• Back view
(Note: I’ve already completed the front-view image generation for this batch. I only need assistance generating photos from other angles.)
(Each original photo set has an uneven number of images and angles, but since there are usually front and back shots, AI should be able to generate photos from various angles.)
(Use AI generation to composite the product onto the model, ensuring consistency between the model, clothing, and product appearance. Each composite image must have a pure white background.)
Resolution:
Each finished image must have a minimum resolution of 1000 × 1000 pixels.
________________________________________
Notes
1. Provided Basic Materials:
A. One photo of the model
B. Four product photos on a mannequin — front, 45°, side, and back views
2. Additional Reference Materials:
A. Reference pose for front view
B. Reference pose for 45° angle view
C. Reference pose for side view
D. Reference pose for back view
(We do not want the AI-generated model to be standing stiffly. The body posture should appear natural and expressive.)
3. Since AI-generated images may include minor product inaccuracies, please cross-check with the original product photos and manually correct any noticeable discrepancies to ensure the final composites match the actual product design.
4. After generation, carefully inspect the following areas for major errors:
a. Metal components – Buckles, clips, or other metal parts must match the real product.
b. Brand rubber label – Ensure the brand mark and its placement are correct.
c. Visible stitching details – Check for correct color, density, and unique cross-stitch patterns.
d. Webbing color errors – Correct any webbing color mismatches to reflect the actual product.
e. Missing functional webbing loops – If the real product includes visible loops or straps that are missing in the AI image, composite them back in.
f. Lighting and shadow consistency – Ensure lighting and shading match naturally across the entire image for realism.
g. Minor visual deviations – Small, non-obvious differences can be ignored; focus mainly on the key visible discrepancies above.
5. If the product’s original photo angles differ from the human-composite angles, use Photoshop to manually adjust and correct the composition perspective.
6. After all adjustments, ensure that the final composite images retain the highest possible original resolution.
AI Multi-Angle Human Model Generation Requirements
Final Deliverable:
Create four AI-generated composite photos of a specified model wearing a specified product, showing the following angles:
• Front view
• 45-degree angle view
• Side view
• Back view
(Note: I’ve already completed the front-view image generation for this batch. I only need assistance generating photos from other angles.)
(Each original photo set has an uneven number of images and angles, but since there are usually front and back shots, AI should be able to generate photos from various angles.)
(Use AI generation to composite the product onto the model, ensuring consistency between the model, clothing, and product appearance. Each composite image must have a pure white background.)
Resolution:
Each finished image must have a minimum resolution of 1000 × 1000 pixels.
________________________________________
Notes
1. Provided Basic Materials:
A. One photo of the model
B. Four product photos on a mannequin — front, 45°, side, and back views
2. Additional Reference Materials:
A. Reference pose for front view
B. Reference pose for 45° angle view
C. Reference pose for side view
D. Reference pose for back view
(We do not want the AI-generated model to be standing stiffly. The body posture should appear natural and expressive.)
3. Since AI-generated images may include minor product inaccuracies, please cross-check with the original product photos and manually correct any noticeable discrepancies to ensure the final composites match the actual product design.
4. After generation, carefully inspect the following areas for major errors:
a. Metal components – Buckles, clips, or other metal parts must match the real product.
b. Brand rubber label – Ensure the brand mark and its placement are correct.
c. Visible stitching details – Check for correct color, density, and unique cross-stitch patterns.
d. Webbing color errors – Correct any webbing color mismatches to reflect the actual product.
e. Missing functional webbing loops – If the real product includes visible loops or straps that are missing in the AI image, composite them back in.
f. Lighting and shadow consistency – Ensure lighting and shading match naturally across the entire image for realism.
g. Minor visual deviations – Small, non-obvious differences can be ignored; focus mainly on the key visible discrepancies above.
5. If the product’s original photo angles differ from the human-composite angles, use Photoshop to manually adjust and correct the composition perspective.
6. After all adjustments, ensure that the final composite images retain the highest possible original resolution.