Number it: Temporal Grounding Videos like Flipping Manga

TL;DR

NumPro introduces a numerical prompting technique that enables Video Large Language Models to perform precise temporal grounding by treating videos as sequences of numbered frames, significantly improving accuracy without extra computational cost.

Contribution

The paper presents NumPro, a novel numerical prompting method that enhances Vid-LLMs' ability to perform temporal localization in videos, achieving state-of-the-art results.

Findings

NumPro improves VTG performance by up to 6.9% mIoU.

NumPro enhances highlight detection by 8.5% mAP.

The method requires no additional computational cost.

Abstract

Video Large Language Models (Vid-LLMs) have made remarkable advancements in comprehending video content for QA dialogue. However, they struggle to extend this visual understanding to tasks requiring precise temporal localization, known as Video Temporal Grounding (VTG). To address this gap, we introduce Number-Prompt (NumPro), a novel method that empowers Vid-LLMs to bridge visual comprehension with temporal grounding by adding unique numerical identifiers to each video frame. Treating a video as a sequence of numbered frame images, NumPro transforms VTG into an intuitive process: flipping through manga panels in sequence. This allows Vid-LLMs to "read" event timelines, accurately linking visual content with corresponding temporal information. Our experiments demonstrate that NumPro significantly boosts VTG performance of top-tier Vid-LLMs without additional computational cost. Furthermore, fine-tuning on a NumPro-enhanced dataset defines a new state-of-the-art for VTG, surpassing previous top-performing methods by up to 6.9\% in mIoU for moment retrieval and 8.5\% in mAP for highlight detection. The code will be available at https://github.com/yongliang-wu/NumPro.