🎬 OmniVDCBench
A bilingual video-domain comprehension benchmark for evaluating fine-grained event understanding, cultural-scene reasoning, and multiple-choice video QA
Dataset Card · Video Question Answering · Fine-grained Event Reasoning · Intangible Cultural Heritage · Chinese / English
Figure 1. Illustration of the OmniVDCBench evaluation setting. The example combines video evidence, multiple-choice questions, model predictions, and category-level… See the full description on the dataset page: https://huggingface.co/datasets/videobenchlab/omniVDCBench.
