Summary of Long Story Short: Story-level Video Understanding From 20k Short Films, by Ridouane Ghermi et al.
Long Story Short: Story-level Video Understanding from 20K Short Films
by Ridouane Ghermi, Xi Wang, Vicky Kalogeiton, Ivan Laptev
First submitted to arxiv on: 14 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed Short-Films 20K (SF20K) dataset is a significant advancement in video understanding, addressing limitations of existing datasets and tasks. The current state-of-the-art models are largely trained on short videos with limited events and narratives, which can be restrictive for real-world applications. SF20K comprises 20,143 amateur films and offers long-term video tasks through multiple-choice and open-ended question answering. Recent vision-language models (VLMs) demonstrate strong performance when fine-tuned on the SF20K-Train set, indicating potential for future progress in long-term video understanding. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper proposes a new movie dataset called Short-Films 20K to help machines better understand videos. Most current datasets are short and show simple actions or scenes with one person. The new dataset has many more movies, some of which are longer and show more complex events. This will help machines learn to understand videos in a more realistic way. The researchers also found that training machines on this new data makes them better at answering questions about long videos. |
Keywords
» Artificial intelligence » Question answering
