We make the first attempt to seamlessly integrate camera geometry into a unified multimodal model, introducing a camera-centric framework, i.e., Puffin, to advance ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback