QEVA: A Reference-Free Evaluation Metric for Narrative Video Summarization with Multimodal Question Answering.

Woojun Jung, Junyeong Kim

Browse the full EMNLP paper archive.