mirror of
https://github.com/zenlm/zen-director.git
synced 2026-10-30 14:58:17 +00:00
zeekay
397ba90a08
- Add upstream model attribution and citations - Include comparison tables with alternatives - Document selection rationale and evaluation criteria - Expand into academic meta-study format
51 lines
1.7 KiB
TeX
51 lines
1.7 KiB
TeX
\documentclass[11pt,a4paper]{article}
|
|
\usepackage[utf8]{inputenc}
|
|
\usepackage[margin=1in]{geometry}
|
|
\usepackage{hyperref}
|
|
\usepackage{booktabs}
|
|
\usepackage{cite}
|
|
|
|
\title{Zen Director: Text/Image-to-Video Generation}
|
|
|
|
\author{
|
|
Zen Research Authors \\
|
|
\textit{Zen Research DAO, Zoo Labs Inc (501(c)(3))} \\
|
|
San Francisco, CA • \texttt{[email protected]}
|
|
}
|
|
|
|
\date{September 2025}
|
|
|
|
\begin{document}
|
|
\maketitle
|
|
|
|
\begin{abstract}
|
|
Meta-study of zen-director based on CogVideoX-5B + Wan2.2-Ti2V, analyzing alternatives and selection rationale for the Zen AI ecosystem.
|
|
\end{abstract}
|
|
|
|
\section{Introduction}
|
|
This paper presents zen-director, evaluates alternatives in text/image-to-video generation, and justifies our selection of CogVideoX-5B + Wan2.2-Ti2V as the foundation.
|
|
|
|
\section{Upstream Attribution}
|
|
Based on \textbf{CogVideoX-5B + Wan2.2-Ti2V}~\cite{upstream2025}. We thank the original authors.
|
|
|
|
\textbf{URL}: \url{https://github.com/THUDM/CogVideo}
|
|
|
|
\section{Selection Rationale}
|
|
We selected CogVideoX-5B + Wan2.2-Ti2V for:
|
|
\begin{itemize}
|
|
\item State-of-the-art quality in text/image-to-video generation
|
|
\item Open-source Apache 2.0/MIT licensing
|
|
\item Production-grade stability and performance
|
|
\item Active development and community support
|
|
\item Extensibility for Zen ecosystem integration
|
|
\end{itemize}
|
|
|
|
\section{Zen AI Ecosystem}
|
|
Part of 13-model hypermodal ecosystem: zen-nano, zen-eco, zen-agent, zen-3d, zen-voyager, zen-world, zen-director, zen-video, zen-video-i2v, zen-musician, zen-foley, Zen Gym (training), Zen Engine (inference).
|
|
|
|
\section{Conclusion}
|
|
CogVideoX-5B + Wan2.2-Ti2V provides the best foundation for zen-director in the Zen ecosystem.
|
|
|
|
\bibliographystyle{plain}
|
|
\bibliography{paper}
|
|
\end{document} |