Stefan Cobeli, Kazi Shahrukh Omar, Rodrigo Valença, Nivan Ferreira, Fabio Miranda

IEEE TVCG · presented at IEEE VIS 2026

The work encodes what can be seen from every point of a 3D city into a compact neural field, making view analyses such as visibility and solar exposure fast enough for interactive exploration. Direct queries compute what is visible from a location; inverse queries find the locations that match a desired view.

← Back