Skip to content

Latest commit

 

History

History
30 lines (18 loc) · 962 Bytes

README.md

File metadata and controls

30 lines (18 loc) · 962 Bytes

Illustration of Pseudo Label Generation

Object Proposals

  • Download our proposals here and put under this folder

    ./proposals/
    ├── coco/
    |   ├── images/
    │       ├── train2017/*.pkl
    
  • each .pkl file contains a list of numpy.ndarray [n_1 * 5, n_2 * 5,...,n_m * 5]

  • the i_th numpy.ndarray correspond to n_i proposals in [xmin, ymin, xmax, ymax, score] of a certain category in the proposal detector

  • each _info.pkl contains image information

Image Caption Dataset

  • We provide an example of image-caption dataset in image_caption_final.json

Object of Interest

  • A train vocabulary contains objects of interest is in object_vocab.json

Download ALBEF Pre-trained Model

  • Download ALBEF pre-trained checkpoint ALBEF.pth and put under this folder