Dual-dialogue digital human (LTX 2.3 Distill 1.1 VBVR 240K)

Motion prompts must be generated using this AI application: https://www.RunningHub.cn/ai-detail/2048722784372662274

Garbled subtitles can be removed using this AI application: https://www.RunningHub.cn/ai-detail/2048760722267578370

This is a high-completion Image-to-Video application based on LTX 2.3 Distill 1.1 + VBVR 240K. After uploading a reference image and filling in the scene description / plot motion prompts, the system generates continuous dynamic video centered around the main subject of the original image. Through three-stage sampling, dual latent upscaling, and I2V conditional control, it further elevates motion execution, detail clarity, and overall completion. The file not only includes LoadImage, PromptRelayEncode, number of frames, and fps, but also chains LTXVImgToVideoConditionOnly, LTXVLatentUpsampler, and CreateVideo into a complete final-cut pipeline. Therefore, it is not an ordinary "single-segment Image-to-Video", but a more professional, final-cut-grade high-quality Image-to-Video workflow.

It is especially suitable for content such as character performances, dual-character dialogues, dramatic short shots, anime character speaking, and mixed 2D and live-action style clips. When using, it is recommended to write clear prompts: who is in the frame, how the positioning is distributed, what action to perform first, how the camera moves, and what state it finally stops in. If you want the motion to be more pronounced, you can increase the motion intensity; if you want the clip to be longer, you can increase the video duration.
🎁Claim RH Coins first, then go experience the workflow! Top-right avatar → Invitation Code → Enter [rh-v1111] to instantly receive 1000 RH Coins, and log in daily to get 100 more coins~ For more ComfyUI workflows, tutorials, and gameplay, welcome to follow our WeChat Official Account (AIKSK), with simultaneous updates on Douyin / Bilibili / Xiaohongshu / YouTube (AI-KSK)✨