Most existing MOT datasets are captured using pinhole cameras, which are characterized by a narrow-FoV and linear sensor motion. However, when panoramic-FoV capture devices experience even slight movements, the entire scene can change drastically, posing significant challenges for object tracking. QuadTrack addresses this challenge by providing a benchmark specifically designed to test MOT algorithms under dynamic, non-linear motion conditions. It enables evaluating algorithm robustness in tracking objects with panoramic, non-uniform motion.
Variants: QuadTrack
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Object Tracking | OmniTrack | Omnidirectional Multi-Object Tracking | 2025-03-06 |
| Object Tracking | DiffMOT | DiffMOT: A Real-time Diffusion-based Multiple … | 2024-03-04 |
| Object Tracking | HybridSORT | Hybrid-SORT: Weak Cues Matter for … | 2023-08-01 |
| Object Tracking | Bot-SORT | BoT-SORT: Robust Associations Multi-Pedestrian Tracking | 2022-06-29 |
| Object Tracking | OC-SORT | Observation-Centric SORT: Rethinking SORT for … | 2022-03-27 |
| Object Tracking | ByteTrack | ByteTrack: Multi-Object Tracking by Associating … | 2021-10-13 |
| Object Tracking | TrackFormer | TrackFormer: Multi-Object Tracking with Transformers | 2021-01-07 |
| Object Tracking | DeepSORT | Simple Online and Realtime Tracking … | 2017-03-21 |
Recent papers with results on this dataset: