Dev uses H3 Max on fal to generate sign language over a movie scene
OdinLovis · x · 2026-10-03
Developer Odin Lovis tried a novel use of the H3 Max video model on fal: generating a video of a movie scene performed in sign language. He admits it's not perfect yet but calls it just the beginning, promising more experiments. A creative demo of video model controllability.
More from Multimodal
- Tuning disocclusion inpainting cadence: smoother output vs hectic detail tradeoff — pixlpa · 2026-10-03
- From one image to an explorable 3D world: img2threejs + Hyper3D + GPT pipeline — Delicious-Shower8401 · 2026-10-03
- One-prompt 15s ad made end-to-end with Opus 5.5 and Hedra MCP in Claude — henloitsjoyce · 2026-10-03
- Constant-size memory for video generative models proposed on Reddit — plsendfast · 2026-10-03
- FreeVideo open-sources local video gen: runs MiniMax H3 on 8GB VRAM laptops — xiuyu_l · 2026-10-03
- Ideogram 4.5 Debuts #23 on Image-Editing Leaderboard, Open Weights Coming — ArtificialAnlys · 2026-10-03