[PaperReading] NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction, NeurIPS 2021 (Spotlight)

# main loop
for iter_i in range(steps):
    # get rays
    # color_fine, s_val, cdf_fine, gradient_error, weight_max, weight_sum = NeuSRenderer.render()
    pass

# render
# uniform sample along rays
# multi pass importance sample along rays