smoke-render

What problem does it solve?

Quickly produce and inspect RGB and depth frames from a robosuite Lift environment to confirm that rendering pipelines and camera setups are producing expected visual outputs before deeper testing.

Core Features & Use Cases

• Multiple renderer support: Run smoke renders using MuJoCo (default) or NVIDIA Isaac Sim.
• RGB and depth capture: Output both RGB and depth PNG frames for visual and quantitative inspection.
• Reporting and display: Locate and display key frames, and report frame dimensions, depth ranges, renderer used, and output location.
• Use cases: Validate renderer integration after environment changes, compare renderer outputs, and perform quick pre-commit or deployment smoke tests.

Quick Start

Run the smoke-render skill with mujoco to produce RGB and depth frames and display agentview_initial_rgb.png and robot0_eye_in_hand_initial_rgb.png from /tmp/render_smoke_frames.