How is the result of Experiment 5.3 (Figure 7) generated?

[leyangjin]

Hello. Nice repo. I have some questions related to the result of Experiment 5.3 (Figure 7). How is these result image generated? Based on the current code, I can't seem to determine where the model checkpoint is located or how it's used to get the final images.

[j0seo]

Hello, the images shown in Fig.7 are the results generated through the reverse process of the diffusion model, equipped with the Consistency Injection Module of 3DFuse (as shown in Fig.3). Currently, the code uses score distillation sampling technique to construct the radiance field for 3D. (This difference is well explained in DreamFusion). Although the code for Fig.7 is not publicly available, you can implement it using the weights provided here.

[leyangjin]

Hello, thank you for your reply. I am not sure whether I understand your response correctly. I wonder whether you mean that the result in Figure 7 is generated just following the process in Figure 3, until the photo after “score distillation” (that is, “4” in the figure below).

Thank you so much for your help and looking forward to your reply.