Splitting Evaluation for "What," "Is," "How," "Can," "Which," and "Others" Sub-Questions

SilongYong / SQA3D

[ICLR 2023] SQA3D for embodied scene understanding and reasoning

Apache License 2.0

117 stars 3 forks source link

Splitting Evaluation for "What," "Is," "How," "Can," "Which," and "Others" Sub-Questions #5

Closed YunzeMan closed 1 year ago

YunzeMan commented 1 year ago

Hi, I noticed that in Table 3 of your paper, you've divided the evaluation into several subsets of questions. I'm curious if you have any corresponding scripts to perform this task. The current codebase appears to only report the overall EM@1 and EM@10 metrics.

jeasinema commented 1 year ago

@SilongYong