vistext
vistext copied to clipboard
VisText is a benchmark dataset for semantically rich chart captioning.
hi, can you list the requirements and package versions for image feature extractor using bottom-up-extractor? I am keep having issues when setting up following the orignal github. I am thinking...
Hi, Thank you for your greate wprk. I was able to train the model with image_guided and datatable input model. Now I am using this trained weights to test on...
Hi, In the `data/statista_mappings.json` file, I see that there is an underscore `_` for each Statista's chart ID from the Chart-to-Text dataset. I have checked the Chart-to-Text's Statista dataset and...