Google Gemini Nano Banana AI saree trend: Common mistakes to avoid while generating your portraits
As part of the viral trend on Instagram, users upload their everyday photographs, which are then transformed into dreamy traditional portraits.
The vintage AI saree edits have taken over Instagram, with users hopping on to the viral trend and generating their own portraits and sharing these.
The users can upload their everyday photographs, which then can transformed into dreamy traditional portraits, which look inspired by a 1990s Bollywood movie.
Created by Google Gemini's Nano Banana tool, these edited images basically picture the users in a saree, providing retro backdrops to the image.
However, many users make some common mistakes while generating these cinematic images, or miss feeding the correct AI prompt, thus resulting in imperfect edits.
Common mistakes to avoid while generating saree portraits
• Using low-quality or group photos - Avoid uploading photographs which are of low quality or a picture with a group of people. AI struggles to create portraits when the face of the subject is not clear. Therefore, make sure that the photograph you upload is a sharp solo picture, where your face is clearly visible and identifiable.
• Providing vague prompts - Be very clear when providing prompts to the AI. For example, vague prompts like “make me look like a Bollywood actor/ actress” will not work. Be specific with the instructions you give. Mention the colour of the saree, its fabric, the lighting style and the background for accurate results.
• Overloading the prompt - Try not to insert too many instructions while feeding the prompt to the AI. Adding too many details can create confusion. Stick to three to four strong and concise visual cues instead of writing long paragraphs.
• Ignoring the background - Specify the background of the edited image. Bollywood posters always have dramatic backdrops, and are sometimes identified with that aesthetic. If you forget to add a cue about the backdrop, the AI might provide a plain default one or use mismatched scenes.
• Forgetting facial consistency - While providing prompts, you can add instructions like “keep the same facial features” and “do not change the face shape”. This prevents the AI from distorting or altering facial features too much.
