Assistance Requested to Locate "Image Feature Volume" and "Aggregated Feature Volume" in GraspNeRF Implementation

[tc-huang]

Hello,

I am currently studying the GraspNeRF paper along with its associated code. As I delve deeper, I am finding it challenging to identify the sections of the code and parameters that correspond to the "Image Feature Volume" and "Aggregated Feature Volume" depicted in Figure 2 (framework of the proposed method) in the paper.

Could someone kindly assist me in pinpointing these components within the code? Any guidance would be greatly appreciated.

Thank you very much for your time and assistance.

Best regards,

tc-huang

[DeepDuke]

I met the same problem. It seems all modules before generating grasp are fused in the NeuralRayRenderer class. The code is too complicated to tell which part is the module in Figure 2 pipeline.