Unverified50% confidenceFactExact time
The experiment uses a DiffusionGemma custom node paired with the LTX 2.3 video generation model, with inputs consisting of one static reference image, one reference video, and one natural language prompt.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/24/2026
First Seen
Valid until: 10/22/2026
Sources
DiffusionGemma角色动作迁移实战：ComfyUI工作流拆解
redditr/comfyui7/12/2026
Related Claims
UnverifiedThe experiment used DiffusionGemma's custom nodes in ComfyUI, paired with the LTX 2.3 video generation model.84% similarUnverified参考视频在技术层面充当「结构先验」角色，纹理细节和光照风格由文本提示词和模型自身预训练知识填充72% similarUnverifiedMidjourney的视频功能采用图像到视频（Image-to-Video）的推理方式，以一张静态图作为条件输入预测后续帧71% similarUnverifiedThe ModelScope text-to-video model is built on a Diffusion Model architecture.71% similarUnverifiedCdance 2.0 supports inputting a single reference image containing multiple scenes, and the model can understand the multiple scenes in the image and convert them into coherent video content.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/607780API
curl https://kongchang.com/api/v1/knowledge/claims/607780MCP
get_claim(id=607780)