PanoPed: Beyond Bounding Boxes for Sim-to-Real Panoramic Pedestrian Tracking

Author1, Author2, Author3, Author4, Author5

angular localizationpedestrian trackingsim-to-realcomputer vision

2610.08826

Builder Relevance

80%

1h ago

Abstract

PanoPed introduces a benchmark for pedestrian tracking using full-sphere panoramic cameras, highlighting the limitations of traditional bounding boxes.

Reality Card

Core Claim

The introduction of the Sextant localization head significantly improves pedestrian tracking performance, raising the HOTA score from 47.30 to 49.49 in the PanoPed-S test.

Method / Result

Sextant improves the MOTIP baseline by 2.19 HOTA points in the PanoPed-S test.

Limitations

The method relies on synthetic data for training and may not generalize well without fine-tuning on real data.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.

← Back to all papers