Spring Semester Requirements
| M.P.1 | Detection | Detect up to 3 pedestrian centroids with Euclidean distance error < 0.3m |
| M.P.2 | Tracking | Track up to 3 pedestrians within 20m of the infrastructure |
| M.P.3 | Prediction | Predict up to 3 pedestrians’ trajectories 1.2 seconds* into the future with an average error of 0.5m |
| M.P.4 | Cycle Time | Time between first frame with a pedestrian to first published trajectory should be less than 0.5 seconds. |
| VEHICLE: | ||
| M.P.5 | Vehicle | Alert driver to stop short of intersection with approaching pedestrian >1.2 seconds* ahead of entry to intersection. |
Spring Validation Experiment Results
SVE Test |
Quantitative Performance |
| Detection Accuracy | 0.12m |
| Tracking Consistency | Continuous tracking |
| Pedestrian trajectory prediction | 0.2745m |
| Cycle Time | 0.24s |
| Vehicle Collision Avoidance Success Rate | 100% |
Use this link to see a video of the final system test from the driver’s point of view.