R-LPIPS Adversarial Training Latency Overhead in Wan2.1 I2V-14B Video Synthesis

This report synthesises findings from 11 peer-reviewed papers addressing the following research question: What is the inference latency overhead introduced by integrating R-LPIPS adversarial robustness training into the Wan2.1 I2V-14B architecture during cross-domain video synthesis. Diffusion-based textit\image-to-video\ (I2V) generation has become a central direction in generative models by turning a reference image, with optional conditions, into a temporally coherent video. Compared with broader video generation settings, this task places stricter. 8 claims were extracted from source literature; 8 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.0/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: What is the inference latency overhead introduced by integrating R-LPIPS adversarial robustness training into the Wan2.1 I2V-14B architecture during cross-domain video synthesis?

Autonomous literature synthesis. Automated review score: 8.0/10. Full text and citation available at Assignee Research.