Outpainting beats OpenPose: a simpler way to control pose in MiniMax H3 videos
bbaudio2024 · reddit · 2026-09-15
A Reddit user shares an alternative approach for controlling character motion in MiniMax H3 videos. The standard fun controlnet route requires converting the control video to OpenPose first, adding processing time and failing when body proportions differ; H3's native reference video mode is slow to sample and often imperfect.
Borrowing the outpainting trick he used with Flux controlnet for consistent characters, the author treats the reference video as an input to outpaint: extend the video, and with the right prompts the extended region naturally becomes the desired result. Verified with his ComfyUI-MMH3-UltimateExtend nodes, it stays consistent even with large proportion mismatches. The trade-off is needing extra video width, which costs more VRAM. Workflow link included.
More from Multimodal
- Midjourney dual --sref combo produces medieval woodcut cat knight art — sergeantsref · 2026-11-02
- Fluffy AI-generated character designs: fox girl, alien bug and cat photographer — OVolosin82152 · 2026-11-02
- Fable 5.5 reportedly one-shots animation in pure code as Anthropic's coding bet pays off — Hesamation · 2026-11-02
- FlowHMR gets 5x faster, demoing an LLM agent pipeline from video to photoreal mocap — camenduru · 2026-11-02
- Retro 1973 music video "Rock is Dead" by TECHHALLA showcases startlingly realistic AI generation — Mother-Pear-2392 · 2026-11-02
- Weekend splatting with COLMAP and Lichtfeld — janusch_patas · 2026-11-02