The concept of Radience Field is used in recent years in deep learning as well. Neural networks are used to approximate the radience field, and the network is called NeRF. Link to the paper: https://arxiv.org/pdf/2003.08934.pdf
curiousdragon
So, to clarify, is the field radiance equivalent to the exiting surface radiance in general? Or are they equivalent only in the case that the given directions that we're evaluating the radiances at are both the surface normal?
I guess another way of putting it would be, does the field radiance only care about the unit area whose normal is the given direction, even when the actual surface has a different normal at that point?
jierui-cell
@longh2000 Thanks for sharing the NeRF paper. It is exciting to see how improved algorithms on field radiance are able to produce high-quality rendering of 3D objects only using several photos from different directions as inputs.
The concept of Radience Field is used in recent years in deep learning as well. Neural networks are used to approximate the radience field, and the network is called NeRF. Link to the paper: https://arxiv.org/pdf/2003.08934.pdf
So, to clarify, is the field radiance equivalent to the exiting surface radiance in general? Or are they equivalent only in the case that the given directions that we're evaluating the radiances at are both the surface normal?
I guess another way of putting it would be, does the field radiance only care about the unit area whose normal is the given direction, even when the actual surface has a different normal at that point?
@longh2000 Thanks for sharing the NeRF paper. It is exciting to see how improved algorithms on field radiance are able to produce high-quality rendering of 3D objects only using several photos from different directions as inputs.