Automated pipeline generates and verifies quiz questions from videos

Automated Video-to-QA Pipeline: End-to-End Question Generation and Verification Using Large Language Models

Computation and Language

Summary

Video content is growing fast, making it hard to find key information and create quizzes. The authors propose an automated system that turns videos into text, summarizes it, and then creates and checks quiz questions using advanced language models. The system improves itself by fine-tuning on verified questions without needing people to do manual labeling. This can help build tools that automatically create study materials from videos.

What this means in practice

  • For online course creators: Automatically generate and verify quiz questions from video lectures to enhance course engagement without manual question writing.
  • For video streaming platforms: Produce structured summaries and assessment questions from vast video collections for improved user knowledge retrieval and learning analytics.

Authors

Jaival Chauhan, Vikram Singh, Pruthwik Mishra

Abstract

The exponential growth of educational and informational video content on platforms like YouTube presents a challenge for efficient knowledge retrieval and assessment. This paper proposes an end-to-end automated pipeline for extracting structured summaries and generating high-fidelity Question-Answer (QA) pairs from video streams. The framework integrates a cascade of state-of-the-art models: audio extraction is handled via FFmpeg, followed by automatic speech recognition (ASR) using Whisper (A library in Python) to generate accurate transcripts. Due to the token constraints of Large Language Models (LLMs), transcripts are segmented and summarized using Facebook's BART. A key innovation of this work is a Self-Verification and Fine-Tuning Loop. We utilize meta-llama/Llama-3.1-8BInstruct to generate candidate QA pairs from transcript chunks, which are then fed back into the model to rigorously assess their relevance and importance. Only verified, high-quality pairs are retained to construct a synthetic dataset. This dataset is subsequently used to fine-tune the Llama model using Low-Rank Adaptation (LoRA), resulting in a specialized model capable of generating domain-aligned assessment questions. This approach demonstrates a scalable method for creating self-improving educational tools without the need for manual data annotation. The proposed model can be downloaded from Kaggle with all the codes from GitHub.