CCN 2025: Incorporating foveal sampling and integration to model 3D shape inferences