Closed shanice-l closed 9 months ago
Only one sentence in 4.1 Evaluation Setup
As the pose distribution in different datasets varies, we organize the manage the test set in a balanced distribution from 0◦ to 30◦.
Thank you for your attention. We will upload the processed dataset to the Hugging Face website.
Thank you for your attention. We will upload the processed dataset to the Hugging Face website.
You still didn't really answer the question, how were these data pairs defined?
Thank you for making the awesome project open-source!
I've noticed that the data pairs are pre-defined in the JSON files (like LMO). So how do you generate these data pairs, by random or according to some principles?
Looking forward to your reply.