In my GLTF scene, I have been exploring the use of the example selection box (code) to select multiple meshes.
Unfortunately, the current approach is providing inaccurate results as it selects based on the centroid of each mesh and includes meshes that are not visible in the camera view (for example, selecting all walls when viewing one side of a house model).
I also tried using the Raycaster, which worked well for mouse/pointer picking.
Is there a way to combine a selection box or two Vector3 points (representing the starting and ending points of the box) with the Raycaster to achieve accurate selections?