SciDuet
Introduced by Sun et al. in D2S: Document-to-Slide Generation Via Query-Based Text Summarization
SciDuet is a dataset for training and benchmarking models for automating document-to-slides generation. It consists of pairs of papers and their corresponding slides decks from recent years' NLP and ML conferences (e.g., ACL). This dataset contains 1,088 papers and 10,034 slides.
Papers
| Paper | Code | Results | Date | Stars |
|---|
