SidewalkBench - Crosswalk (Procedural): leaderboard
Navigation policies drive a simulated delivery robot 10-20 m across obstacle-free crosswalks to the kerb ramp in 100 procedurally generated scenarios without pedestrians; the number is route completion in percent, mean of 5 seeded runs. It isolates lane-following and obstacle avoidance; the pedestrian and long-route boards add people and distance.
Metric: Route completion on crosswalk blocks in procedurally generated scenes, 100 scenarios, mean of 5 seeded runs (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MIMIC | 98.8 | |
| 2 | FlowPilot | 98.3 | |
| 3 | S2E | 95.9 | |
| 4 | MBRA | 95.5 | |
| 5 | NoMaD | 94.4 | |
| 6 | CityWalker | 77.4 | |
| 7 | InternVLA-N1-DualVLN | 72.8 | |
| 8 | ViNT | 30.9 |
Interactive version: theaggregate.ai/benchmark?slug=sidewalkbench-crosswalk-procedural · How It Works · Data refreshed daily, snapshot 2026-11-13.