Research Trends in AI-Driven Video Production A Systematic Bibliometric Investigation
TL;DR
The intellectual evolution of this field from 2016 to 2026 is mapped, focusing on key publication trends and thematic shifts, indicating that the foundational frameworks of the field were established during that time.
Abstract
Artificial intelligence (AI) has fundamentally reshaped the world of video production, moving from simple automated edits to complex generative models capable of creating realistic content from scratch (Xie et al., 2025). This study maps the intellectual evolution of this field from 2016 to 2026, focusing on key publication trends and thematic shifts. A systematic bibliometric study of 2,371 peer-reviewed papers included in the Scopus database was carried out. The metrics included the total number of publications (TP), total citations (TC), the number of cited publications (NCP), average citations per publication (C/P), and average citations per cited publication (C/CP). The data reveals a massive surge in research output starting in 2021, with total publications reaching a peak of 796 in 2025. While volume is currently at an all-time high, the highest citation impact per paper occurred in 2019, indicating that the foundational frameworks of the field were established during that time. Thematically, research is clustered into four main areas: core technical architectures, media synthesis, deepfake detection and human ethics. This investigation provides a comprehensive roadmap for researchers navigating the fast growth of AI-driven media production.
