MARSAD / CONSOLIDATED CATALOG / #014

Headphones and an unresponsive straight crossing

Awaiting your 3D scene approval

A headphone-wearing pedestrian walks straight into the lane without reacting to a warning, then completes the crossing.

Original description: 前方有行人佩戴着降噪耳机,听不到车辆靠近的警示音,径直走入车道。

47 sFull continuous rehearsal
2Critical actors and props
941Exported time steps checked
0New Image 2.5 outputs

Loading the full animated scene…

Critical moments

What to verify

  • Verify exactly one adult and one persistent headphone prop, with no phone, extra pedestrian or detached duplicate headset.
  • Evaluate 941 samples at 20 Hz over -7..40 s; keep x=43.2 throughout and maintain grounded walking with bounded acceleration and deceleration.
  • Show no warning-related head turn, evasive step or lane pause; this must remain distinct from the phone-distraction scene.
  • Verify the complete adult footprint is within the south sidewalk initially and within the north sidewalk at 8.8 s; check existing curb transitions without adding a ramp.
  • Keep ego stopped at x=12, y=0 until at least 11 s and release only after the pedestrian's full articulated envelope remains clear.
  • Require human approval of the entire rehearsal before Image 2.5 edits original Image 4; use the rehearsal for position and posture, not replacement architecture.

Review notes

Reviewed the full straight crossing, headphones and steady walking. The headset remains attached and facing follows the crossing path. Contact inspection was reframed from the road side.

Check action, relative position, facing, posture and visibility. Approximate offsets are acceptable. Sampled visual checks do not prove every contact or continuous collision clearance.

Approval stops here. Please identify this scene as #014 when approving it or requesting revisions. Image 2.5 generation and training remain paused for this new catalog phase.

Rendered inspection views

Event, approach and final views rendered from the exact committed Higgsfield Blender source. These are procedural 3D previews, not generated training photographs.

Original NAVSIM camera context

4 conflicting background vehicles were moved to clear holding positions for this counterfactual rehearsal. Original photos remain unchanged. Before later inpainting, any overlap with original-photo traffic must be resolved. Background positions in this 3D rehearsal are not an exact replay of the photographs.

After scene approval

The working pipeline will edit each original NAVSIM camera photograph with Image 2.5. Its same-time Higgsfield view will provide position and posture, annotated with facing direction and current speed. Temporal image requests will cover the full event with a 1 Hz base plus critical moments.

Assumptions and source provenance

Fresh GPT-6 motion plans use semantic descriptions from the pinned consolidated catalog and NAVSIM environments. No upstream nuPlan vectors, trajectories or images are used. The four source environments are reused; 82 scenario descriptions do not mean 82 independent locations.

Higgsfield revision 1. The portable GLB combines export clips without changing geometry or animation samples.

This review uses Higgsfield revision 1 and the exact exported scene snapshot.

Higgsfield workspace (latest revision) · Download animated GLB · GPT-6 scene plan · Visual review evidence · Motion checks · Approval status