Single-illustration real-time anime character demo built on See-through and Anime2.5DRig
bdsqlsz · x · 2026-10-27
An indie team unveiled their mascot "若泉绮羽": from a single character illustration, the character blinks, turns, changes expressions and responds to pokes — all driven in real time with no hand-keyed animation, voiced by GPT-SoVITS.
Stack:
- See-through (shitagaki-lab, SIGGRAPH 2026 paper, 4.9k GitHub stars): single-image layer decomposition for anime characters.
- Anime2.5DRig (8co28): animates the decomposed layers in the browser.
The team heavily reworked both open-source projects so no rigging skills or manual layer splitting are needed. The project will be open-sourced soon; half-body effects shown first.
More from Multimodal
- Full prompt revealed for AI-generated real estate tour videos with timestamped shots and lip sync — FinanceYF5 · 2026-10-27
- First working AI real estate video workflow: Kling 4 one-shot with consistent agent and house — FinanceYF5 · 2026-10-27
- 40-person blind test: people spot AI video at 52%, barely better than a coin flip — Resident-Anxiety6427 · 2026-10-27
- China's first AI feature film flops at the Chinese box office — kernelangus420 · 2026-10-27
- Dessa full-duplex speech-to-speech model runs on consumer GPUs, dubbed a 'GPT-3 moment' for speech — Puzzleheaded-King584 · 2026-10-27
- Redditor shares eerie dreamlike AI video of a dim attic bedroom — Sweaty-Ad4488 · 2026-10-27