facebookresearch / mmf

A modular framework for vision & language multimodal research from Facebook AI Research (FAIR)
https://mmf.sh/
Other
5.46k stars 932 forks source link

Which model is used to get trainval2014.lmdb? #1208

Open ziyanyang opened 2 years ago

ziyanyang commented 2 years ago

❓ Questions and Help

Hi,

I have a question about the visual bert code. If I want to evaluate some tasks using VisualBert pre-trained on COCO, I guess I need to reuse the features from trainval2014.lmdb. But If I want to get these features for a new image, which file should I focus on? I saw https://github.com/facebookresearch/mmf/blob/main/tools/scripts/features/frcnn/extract_features_frcnn.py and https://github.com/facebookresearch/mmf/blob/main/tools/scripts/features/extract_features_vmb.py. What is the difference between them? And how can I get the same features for trainval2014.lmdb?

Thanks!