Question about the paper and implementation (IC, and CC evaluation)

[jihoontack]

Hi, thank you for the awesome work, also the implementations are awesome!

While reading your paper, I have seen you evaluated the ShapeNet dataset for both intra-category (IC) and cross-category (CC) levels.

However, it was hard for me to find the options in the implementation.
How can I evaluate IC and CC on this implementation?

Thank you again for the awesome work!

Best,
Jihoon

[gaobaoding]

Hi @jihoontack,

Thanks for your interest in our work. Evaluations on both IC and CC are executed automatically using

python evaluation.py --config cfg/evaluation/${config file}

which will then run the code.
Also, you can use a for-loop to evaluate the performance with different number of contexts as shown in the paper.

Best,
Ning

[jihoontack]

Thank you very much for the answer!

Really enjoyed reading your paper