Tuning disocclusion inpainting cadence: smoother output vs hectic detail tradeoff
pixlpa · x · 2026-10-03
The author is experimenting with different disocclusion inpainting and regeneration cadences in AI video generation, balancing smear trails against a jarring rhythmic element. Finding: more frequent inpaints are less obvious but lead to more hectic, busy detail.
More from Multimodal
- Micro-movie 'MOTHER' Explores Post-Human Androids Stuck in Pointless Routines — Kyrannio · 2026-10-03
- Jordan Chesney's AI Filmmaking Livestream Called 'Absolute Gold' for Creators — Kyrannio · 2026-10-03
- Creator shares prompts for dropping your performance into any world with Luma AI — mrjonfinger · 2026-10-03
- How an AI-made Bach podcast was built with Gemini 3.8 Flash and heavy human steering — sytelus · 2026-10-03
- Dev uses AI to build a free podcast that unpacks Bach measure by measure — sytelus · 2026-10-03
- From one image to an explorable 3D world: img2threejs + Hyper3D + GPT pipeline — Delicious-Shower8401 · 2026-10-03