Text-to-Edit: Controllable End-to-End Video Ad Creation via Multimodal LLMs

The exponential growth of short-video content has ignited a surge in the necessity for efficient, automated solutions to video editing, with challenges arising from the need to understand videos and tailor the editing according to user requirements. Addressing this need, we propose an innovative end-to-end foundational framework, ultimately actualizing precise control over the final video content editing. Leveraging the flexibility and generalizability of Multimodal Large Language Models (MLLMs), we defined clear input-output mappings for efficient video creation. To bolster the model's capability in processing and comprehending video content, we introduce a strategic combination of a denser frame rate and a slow-fast processing technique, significantly enhancing the extraction and understanding of both temporal and spatial video information. Furthermore, we introduce a text-to-edit mechanism that allows users to achieve desired video outcomes through textual input, thereby enhancing the quality and controllability of the edited videos. Through comprehensive experimentation, our method has not only showcased significant effectiveness within advertising datasets, but also yields universally applicable conclusions on public datasets.

• We propose a novel video editing framework with a multimodal LLM that clearly defines input-output formats, streamlining production by managing material comprehension and arrangement in one step. To the best of our knowledge, this is the first end-to-end MLLM-based video editing framework.
• We introduce an effective approach employing a denser frame rate and a slow-fast processing strategy, which significantly enhances the model's ability to extract and understand temporal and spatial video information.
• To improve the controllability of the model's output, we develop a text-driven video editing method that ensures the final output aligns precisely with user expectations.
• We conduct comprehensive experiments and exploratory analyses, demonstrating that our proposed video editing framework can optimally perform advertising short-video editing tasks. Moreover, our framework generalizes well to public datasets, achieving applicable results.