AI is good at this. Although you can't expect it to be very detailed, AI always simplifies and often makes things smaller and worse. You need to make several attempts and understand what he offers. Agree on the best option. For example, these are the flowers in the photo and this is the result after creating the model
another bouquet
still
It can be seen that AI has simplified all the photos very much. But in the end, you can get a variety of bouquet options. I took the finished model of the vase, added 0.2 and Ni 1.5 (optical refraction coefficient of light in glass) to the MTL file. I made water out of a cylinder. First, I assigned it a width and depth like a vase, then I leveled everything and after that I reduced all the dimensions of the cylinder by 1 cm. The water rose evenly inside the vase. Powerful small blue spheres of 10% each should be placed inside the water. they will provide the necessary light spots from the water next to the vase. To create models from photos, I used TRELLIS, improved the rendering quality using REMINI.
