search

tuyunbin / SCORER

[ICCV 2023] This is the Pytorch code for our paper "Self-Supervised Cross-View Representation Reconstruction for Change Captioning".
14 stars 1 forks source link

Could you post transformer_vocab.json and transformer_labels.h5 files? #9

Open wenting666 opened 2 months ago

wenting666 commented 2 months ago

FileNotFoundError: [Errno 2] No such file or directory: './data/transformer_vocab.json'

tuyunbin commented 2 months ago

Hi, you can obtain these files by this command:

Build vocab and label files using caption annotations:

python scripts/preprocess_captions_transformer.py --input_captions_json ./data/change_captions.json --input_neg_captions_json ./data/no_change_captions.json --input_image_dir ./data/images --split_json ./data/splits.json --output_vocab_json ./data/transformer_vocab.json --output_h5 ./data/transformer_labels.h5