# shitagaki-lab/see-through

> A research project that decomposes a single anime illustration into semantic layers (up to 23: hair, face, eyes, clothing, etc.) with inpainting of occluded regions and inferred drawing order, exporting the result as a PSD.

- Magnitude: 3.6 out of 10 — Breakout
- Stars: 5,083 total · +21 stars measured 2026-10-28, 11:23–15:57 UTC, ≈ 39 by evening
- Star trust: star growth looks organic
- Category: Generative media · Language: Python · License: Apache-2.0 · Created: 2026-03-31 · Last push: 2026-10-05
- GitHub: https://github.com/shitagaki-lab/see-through · Page: https://gitnova.dev/en/r/shitagaki-lab/see-through

## Useful for

- Decompose an anime illustration into PSD layers for later animation or character rigging
- Split a character by depth or left-right via heuristic_partseg for targeted edits
- Run the pipeline on 8-12 GB VRAM using group_offload or NF4 quantization

## Why it’s here

- Star-counter measurements on 2026-10-28 (UTC), 11:23–15:57: 5062 → 5083 stars (+21). This is the change over that interval.
- Estimated end-of-day forecast: about +39 stars, using observed gains and the previous day.
- Over the last two days the pace is 3.7× that of the previous week and a half.
- Elevated gains were measured on the last 2 completed UTC days in a row. Continued growth is not guaranteed.
- GitHub Trending Python today: #9, +78 stars.
- About 7 forks a day — people are taking the code.

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 448
- Issues and pull requests: 46
- Watchers: 38
- Average over the last week: 46 per day
- Usual pace: 24 per day
- Stars in the last hour (measured): 1

## Stars per day, last 30 days (oldest → newest, today is partial)

2026-09-29 … 2026-10-28: 15, 15, 15, 27, 79, 30, 38, 26, 32, 34, 39, 11, 22, 23, 35, 25, 17, 15, 24, 26, 18, 22, 26, 19, 8, 19, 23, 106, 106, 21

## Spotted in now

- GitHub Trending Python today: #9, +78 stars

## More in this category

1. **anhao/ai-pose-gallery-zh** — 8.6 · Breakout · Generative media · HTML · +1,310 stars measured 2026-10-28, 00:02–15:55 UTC, ≈ 1,841 by evening
   Chinese localization of an AI pose reference gallery: 1439 records and 2878 images with a ready-to-deploy dependency-free static site. Useful for artists and 3D modelers picking poses.
   Full card: https://gitnova.dev/en/r/anhao/ai-pose-gallery-zh.md
2. **toyxyz/toyxyz_RGBMocap** — 5.8 · Early signal · Generative media · Python · +91 stars measured 2026-10-28, 00:02–15:55 UTC, ≈ 137 by evening
   Native Windows app for video motion capture: reconstructs 3D human animation with ExMR, FlowHMR and GVHMR models, corrects contacts and exports FBX/VMD.
   Full card: https://gitnova.dev/en/r/toyxyz/toyxyz_RGBMocap.md
3. **debpalash/VoiceStudio** — 5.4 · Steady · Generative media · Python · +1,278 stars measured 2026-10-28, 00:02–15:56 UTC, ≈ 1,906 by evening
   A fully local, open-source ElevenLabs alternative for voice cloning, voice design, video dubbing, dictation, transcription and audiobook creation on your own hardware, with no account or API key. Supports 16 TTS and 11 ASR engines, a…
   Full card: https://gitnova.dev/en/r/debpalash/VoiceStudio.md
4. **0xS4N4TG/kinetograph** — 5.1 · Early signal · Generative media · TypeScript · +98 stars measured 2026-10-28, 00:03–15:56 UTC, ≈ 142 by evening
   A programmatic video engine in TypeScript with a Rust GPU compositor: renders video from code, supporting Remotion API, CSS, SVG, Canvas 2D, WebGL, and Web Audio.
   Full card: https://gitnova.dev/en/r/0xS4N4TG/kinetograph.md
5. **Tencent-Hunyuan/Hy-Nomas** — 4.8 · Early signal · Generative media · Python · +78 stars measured 2026-10-28, 00:03–15:56 UTC, ≈ 114 by evening
   An interactive video world generation model that creates video segments from free-form text prompts or camera trajectories using autoregressive generation with a lifelong KV cache. Includes inference, LoRA fine-tuning, data annotation,…
   Full card: https://gitnova.dev/en/r/Tencent-Hunyuan/Hy-Nomas.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-28 16:10 UTC, updated every 30 minutes.