Hi, thanks for your great work!
The code in sdfusion_model#399 is used to complete a partial shape, but this code: sdfusion_model#417 uses full shape to encode latent z and uses mask to get partial z, and then complete the shape.
But if I want to complete a partial shape, I don't have a full shape to encode, should I just use the partial shape to encode and continue the next steps?
Can you explain the completion process? Thx!
Hi, thanks for your great work! The code in sdfusion_model#399 is used to complete a partial shape, but this code: sdfusion_model#417 uses full shape to encode latent z and uses mask to get partial z, and then complete the shape. But if I want to complete a partial shape, I don't have a full shape to encode, should I just use the partial shape to encode and continue the next steps? Can you explain the completion process? Thx!