google-deepmind/robotics-antigravity-plugin

mujoco-showcase-shot

Renders a studio-lit 360-degree turntable showcase video (.mp4) and keyframe sheet for a MuJoCo asset (MJCF model.xml, .mjz, or mjSpec/MjModel) using MuJoCo's Filament PBR renderer (mjrf), sweeping articulated joints, pulsing collision geoms, and drawing a…

ソースを見る

使い方

  1. インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
  2. 必要なツール、認証情報、対応環境を原文で確認します。
  3. 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/google-deepmind/robotics-antigravity-plugin --skill mujoco-showcase-shot

開始前の確認

  • 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
  • スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
リポジトリの原文

見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。

MuJoCo Filament Showcase Shot Turntable Video Generator

Use this skill whenever you want to generate a polished, studio-lit 360° turntable showcase video (`.mp4`) for a MuJoCo MJCF asset (model.xml) using MuJoCo's Filament PBR renderer (`mjrf`).

What It Showcases

  1. **Filament Studio Softbox IBL, Gradient Skybox & Reflective Turntable

Plinth**:

  • Automatically wraps the asset in a studio scene with a procedural

6-face studio softbox cubemap IBL (studio_ibl_tex) paired with a balanced 4-point physical studio lighting rig (shadow-casting key spotlight, cool fill, warm rim, and upward floor-bounce directional light), ACES tone mapping, high-quality SSAO, and subtle bloom.

  • Places the asset on a smooth 96-slice two-tier chamfered turntable

plinth with a brushed metallic accent ring resting on an infinite reflective studio checker floor (reflectance="0.08").

  1. 2-Orbit 360° Turntable Sequence + Articulation Playback:
  • Completes Orbit 1 (`0%–50%`, full 360°) showing pure textured visual

meshes so users can inspect every side uninterrupted, followed by Orbit 2 (`50%–100%`, second 360°) cycling through physics collision verification while smoothly sweeping all <joint type="hinge"/> and <joint type="slide"/> degrees of freedom via a cosine trajectory.

  1. Cross-Dissolved Physics Collision Geometry Inspection:
  • Transitions via a smooth alpha cross-fade (~8–10 frames, avoiding hard

1-frame pop-ins) across three plain-English chapters:

  • `0%–50%` (Orbit 1 — Full 360°): Textured Visual (`group="2"`)
  • `50%–78%` (Orbit 2a): **Visual + Translucent Ghost Collider

Overlay** (uniform mint-cyan ghost shell over the textured mesh)

  • `78%–100%` (Orbit 2b): Physics Collision Hulls Only (visual

mesh hidden; per-hull color-coded convex shapes)

  1. Self-Explanatory HUD Overlay:
  • Top-left translucent card showing Asset Name, **Bounding Box Size

(L × W × H in mm / cm), Total Mass (g / kg), Visual / Collider counts, and Live Joint Angles/Positions**.

  • Top-right chapter badge (1/3 · TEXTURED VISUAL (360°) /

2/3 · VISUAL + COLLIDER OVERLAY / 3/3 · COLLISION HULLS ONLY) and orbit angle readout (Orbit 1/2 / Orbit 2/2).

  • Bottom-center plain-English mode callout banner and **3-segment chapter

timeline bar** showing all three phases at a glance.


Command-Line Usage

Resolve SKILL_DIR to the directory containing this SKILL.md file. Bundled `scripts/render_showcase_shot.py` includes a PEP 723 # /// script header (mujoco, av, numpy, Pillow, absl-py), so uv run automatically resolves and caches an isolated environment on first use (ensure ffmpeg is also installed when passing --before_xml). Always write output videos and images to <ArtifactDir>/ (never inside a source repository):

bash
SKILL_DIR="/path/to/skills/mujoco-showcase-shot"
uv run "$SKILL_DIR/scripts/render_showcase_shot.py" \
  --asset=/path/to/model.xml \
  --output=<ArtifactDir>/showcase_shots/<asset_name>_turntable.mp4 \
  --width=960 \
  --height=720 \
  --fps=30 \
  --duration=8.0

This produces both:

  • <ArtifactDir>/showcase_shots/<asset_name>_turntable.mp4 (H.264 turntable

video)

  • <ArtifactDir>/showcase_shots/<asset_name>_turntable_keyframes.png (2×2

keyframe summary image across the 4 showcase phases)

When imported from Python (import render_showcase_shot after adding $SKILL_DIR/scripts to sys.path), this module also provides:

  • render_6view_inspection_sheet: Renders a 4×7 multi-view inspection sheet

(Front (+X), Back (-X), Right (-Y), Left (+Y), Top (+Z), Bottom (-Z), and 3/4 Front-Right) across PBR Visual (group="2"), Untextured Clay 3D Form (group="4"), Visual + Collision Overlay (group="2" + "3"), and Per-Hull Collision (group="3") modes.

  • render_side_by_side_comparison: Renders Reference vs. Filament PBR vs.

Untextured Clay comparison sheets (including clay_only=True for fast early geometry iteration, detail_crops_out_png for 1:1 high-resolution zoom crops, and six_view_out_png for single-pass multi-view output).

  • render_wireframe_and_uv_sheet: Renders a 2×3 raw .obj quad/triangle

topology wireframe and per-material UV layout inspection sheet.

  • render_filament_still: Renders batched Filament PBR, Clay, Overlay, or

Collision still views.

同じリポジトリから

関連する Skills

すべての Skills
google-deepmind
コミュニティ

blender-creation

Creates physically calibrated, articulated, PBR-textured 3D MuJoCo assets (MJCF model.xml, .mjz, mjSpec) from scratch using headless Blender (bpy, bmesh, blendermeshutils, assetbuildutils), convex collision synthesis, 10-scenario auditing, and Studio preview. Use only when asked to create, model, or build a 3D MuJoCo object from a URL, image, or text description, or when blender-creation is named. Don't use for non-MuJoCo Blender/Cycles/CAD modeling.

1時間の導入数
2
GitHub Stars
80
更新日
11月17日
google-deepmind
コミュニティ

fr3-duo-control

Franka Duo robot only (Franka Research 3 / FR3 Duo dual-arm cell with Robotiq 2F-85 grippers, ZED Mini and RealSense D405 cameras). Brings up, homes, and controls it via pylibfranka on hardware or in MuJoCo sim. Use only when the user names Franka Duo, FR3 Duo, or fr3duo.

1時間の導入数
2
GitHub Stars
80
更新日
11月17日
google-deepmind
コミュニティ

i2rt-yam-bringup

Checks and installs the I2RT YAM SDK (i2rt), configures Linux socketcan or macOS gsusb CAN bringup for I2RT YAM robot arms, and provides a wave smoke test (wave.py). Use only before bringing up or programming a physical or simulated I2RT YAM robot arm, or when asked to run a wave gesture on an I2RT YAM robot. Don't use for non-YAM CAN bus debugging (e.g. automotive OBD-II or industrial PLC CAN networks) or non-robotics tasks.

1時間の導入数
2
GitHub Stars
80
更新日
11月17日
google-deepmind
コミュニティ

i2rt-yam-ik

Solves 6-DOF inverse kinematics for the I2RT YAM and YAM Pro robot arm using Mink + DAQP on combined YAM + gripper MJCF. Use only when computing joint targets or local Cartesian moves on an I2RT YAM or YAM Pro robot arm. Don't use for non-YAM inverse kinematics, general quadratic programming (QP) solvers, or non-robotics tasks.

1時間の導入数
2
GitHub Stars
80
更新日
11月17日