ML Algorithm which reconstruction of 3d models using multi-view images of the product, but specifically looking train model which reconstructs 3D models of trainers.
- The input images can be feed in random order and it can take one or more images of an object instance from arbitrary viewpoints and outputs a reconstruction of the object in the form of a 3D. Examples of the input image [login to view URL], We also have images of trainers which are with up to 30+ perspective images.
- The model must also include the trainer texture which is extracted from the picture and applied to the 3D models. The texture should be high quality as this would use again one-commence sites and will be integrated with AR.
- preferred output file (obj).
- Must have command line interface or web interface, command line interface ie. [[login to view URL]]([login to view URL]) -img /path_to_trainer_imagesOfATriner -o output of the 3D Model
- Then this will be stored in a database or local file storage.
I don't want anyone inexperienced making any offers if read this proposal and interested to discuss this further or any question, please reach out. Please include this ref in the comment: 23432423423.