Skip to content
/ CoVS Public

Coherent Visual Storytelling via Parallel Top-Down Visual and Topic Attention

Notifications You must be signed in to change notification settings

JinjingGu/CoVS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CoVS

Source code of CoVS for Visual Storytelling task.

Environments

  • CUDA 10.1
  • python 3.6
  • pytorch 1.6.0

Run

Datasets

Download VIST ResNet152 features and put in project directory, make sure in src_xxx/dataset.py the path is correct. (Acutally you can generate ResNet152 features from original dataset)

Generate Rake or LDA topics

cd src_xxx
python dataset.py

More details in dataset.py. For example, you need to call VISTDataset.init_keywords_lda() to generate LDA topics.

Train or Test model

cd src_xxx
python main.py

You can change options(including 'train' or 'test') and hyperparameters at src_xxx/main.py.

Acknowledgement

License

MIT

About

Coherent Visual Storytelling via Parallel Top-Down Visual and Topic Attention

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages