Abstract
Lecture slide element detection and retrieval are key problems in slide understanding. Training effective models for these tasks often depends on extensive manual annotation. However, annotating large volumes of lecture slides for supervised training is labor intensive and requires domain expertise. To address this, we propose a large language model (LLM)-guided synthetic lecture slide generation pipeline, SynLecSlideGen, which produces high-quality, coherent and realistic slides. We also create an evaluation benchmark, namely RealSlide by manually annotating 1,050 real lecture slides. To assess the utility of our synthetic slides, we perform few-shot transfer learning on real data using models pre-trained on them. Experimental results show that few-shot transfer learning with pretraining on synthetic slides significantly improves performance compared to training only on real data. This demonstrates that synthetic data can effectively compensate for limited labeled lecture slides. The code and resources of our work are publicly available on our project website: this https URL.
Abstract (translated)
讲义幻灯片元素检测和检索是理解幻灯片的关键问题。为了有效地训练用于这些任务的模型，通常需要大量的手动注释。然而，对大量讲座幻灯片进行监督训练的手动标注既耗时又需要领域专业知识。为了解决这个问题，我们提出了一种由大语言模型（LLM）引导的合成讲座幻灯片生成管道SynLecSlideGen，该管道可以产生高质量、连贯且真实的幻灯片。此外，我们创建了一个评估基准——RealSlide，通过手动注释1,050张真实讲座幻灯片来构建。为了评估我们的合成幻灯片的实用性，我们在真实数据上进行了少量样本迁移学习，并使用在合成幻灯片上预训练的模型进行测试。实验结果表明，与仅在真实数据上训练相比，在合成幻灯片上进行预训练后的少量样本迁移学习显著提高了性能。这证明了合成数据可以有效地弥补标记讲座幻灯片数量不足的问题。我们的工作代码和资源可以在项目网站上公开获取：[此URL]。
URL
https://arxiv.org/abs/2506.23605
