Zephyr Generative Pre Trained Photogrammetry

This was our April’s fools 2023 joke 🙂 we hope you liked it!

We are happy to announce our new project – deep learning based photogrammetry!

We have trained a model called ZGPP (Zephyr Generative Pre-trained Photogrammetry) which interacts in a conversational way. The dialogue format makes it possible for ZGPP to dynamically generate 3D models, answer followup questions, admit its mistakes and challenge incorrect premises.

ZGPP is a state of the art SLP* (single layer perceptron neural network) trained to generate the best possible models, both in terms of geometry and textures.

You can access this artificial intelligence state-of-the-art functionality for free, right here, right now!

Just type a question (aimed to the generation of a 3D model – please keep in mind this is not a general purpose chatbot) and ZGPP will dynamically create a fully feature textured mesh, in seconds!

Please note that at this time ZGPP only speaks English.

Example: Hello ZGPP! Can you create a 3D model of a cherub statue for me?

Access ZGPP – 100% FREE!

ZGPP : Hello, I am ZGPP. What 3D model would you like me to generate for you today?

*The training dataset is composed by 1 dataset, “Cherub Statue” by 3Dflow SRL.