search

aimagelab / meshed-memory-transformer

Meshed-Memory Transformer for Image Captioning. CVPR 2020
BSD 3-Clause "New" or "Revised" License
507 stars 136 forks source link

I don't know what 'cls_prob' means #70

Open yoojin9649 opened 2 years ago

yoojin9649 commented 2 years ago

Hello. Thanks for sharing your code.

I checked coco_detections.hdf5 file using HDFView.

I understand that imgid_features and imgid_boxes are extracted from Faster R-CNN(with ResNet101).

But, I was wondering how you extracted the cls_prob and what it means.

Thanks for reading!

cxy990729 commented 2 years ago

HI, I have the same problem. Do you know how to solve it now？ @yoojin9649

songbohui commented 10 months ago

Hello. Thanks for sharing your code.

I checked coco_detections.hdf5 file using HDFView.

I understand that imgid_features and imgid_boxes are extracted from Faster R-CNN(with ResNet101).

But, I was wondering how you extracted the cls_prob and what it means.

Thanks for reading!

Hello, excuse me, I recently read the paper Meshed_memory, and I am a little confused about the data set coco_detections.hdf5 he used. Is the value extracted by the imgid_features key the regional feature of the image corresponding to the image_id? Or can you tell me what content this imgid_features is used to save?