Hi,
I’m experiencing frequent content violations with the DALL-E 3 API, even though I use the Moderation API to check my prompts beforehand. The Moderation API consistently marks my content as safe, indicating no content violations. My content is primarily related to developmental milestones of young children and expecting mothers, emphasizing positive and educational themes.
Example Prompts:
- A photorealistic close-up of a happy, chubby baby’s knees as they crawl on a colorful play mat. The image highlights the soft, squishy texture of the knees, with a simple, blurred background in soft pastel colors. The image is designed to be visually appealing and attention-grabbing for a thumbnail.
- A pregnant woman in athletic wear performing yoga in a serene park setting. She looks joyful and energized. In the background, other pregnant women are engaged in different safe exercises such as walking, light jogging, and swimming in a nearby pool. The sky is clear and blue, and the overall atmosphere exudes health, positivity, and well-being. The woman’s baby bump is visible, symbolizing the benefits of staying active during pregnancy.
- Create a photorealistic image featuring an adorable infant crawling with soft, pudgy knees. Highlight the baby’s knees with subtle emphasis to show their flexibility and cartilaginous nature. The background should be a cozy, home environment to indicate a safe and nurturing space. The infant should appear happy and curious, with the focus on the baby’s knee.
- Create an image featuring a mother and her 2-month-old baby, engaged in a playful learning activity. The baby is lying on his back on a playmat, and the mother is gently touching the baby’s tiny toes while smiling and singing. The setting is a cozy nursery room with soft colors, emphasizing the bonding and developmental benefits of this activity.
This violation error is significantly affecting my work. This issue arises only with the DALL-E 3 API; ChatGPT’s DALL-E integration does not give any violations for the same prompts. I would appreciate any insights or guidance on why this might be happening and how to resolve the issue.
Note: I have already added sentences like: ‘The image is intended for educational purposes. Avoid any content that may be considered inappropriate or offensive, ensuring the image aligns with content policies. Adjust the prompt to eliminate any content violations if needed.’ in my prompt to avoid content violation policy, but nothing works.
