Structure Preserving Object Tracking
Model-free trackers can track arbitrary objects based on a single (bounding-box) annotation of the object. Whilst the performance of model-free trackers has recently improved significantly, simultaneously tracking multiple objects with similar appearance remains very hard. In this paper, we propose a new multi-object model-free tracker (based on tracking-by-detection) that resolves this problem by incorporating spatial constraints between the objects. The spatial constraints are learned along with the object detectors using an online structured SVM algorithm. The experimental evaluation of our structure-preserving object tracker (SPOT) reveals significant performance improvements in multi-object tracking. We also show that SPOT can improve the performance of single-object trackers by simultaneously tracking different parts of the object.
Code (0)
등록된 구현이 없습니다.
Tasks
Multi-Object TrackingObjectObject TrackingSimilar Papers 제목 키워드 기반
Structure From Tracking: Distilling Structure-Preserving Motion for Video Generation
Reality is a dance between rigid constraints and deformable structures. For video models, that means generating motion that preserves fidelity as well as structure. Despite progress in diffusion models, producing realist…
Video GenerationObject Preserving Siamese Network for Single Object Tracking on Point Clouds
Obviously, the object is the key factor of the 3D single object tracking (SOT) task. However, previous Siamese-based trackers overlook the negative effects brought by randomly dropped object points during backbone sampli…
3D Single Object TrackingObjectObject LocalizationObject TrackingConTrack: Constrained Hand Motion Tracking with Adaptive Trade-off Control
Human demonstrations provide strong priors for robot manipulation, yet it is non-trivial to transfer them to execute on real robots due to the kinematic gap. In dexterous manipulation, it remains challenging to track lon…
Reinforcement LearningRobot ManipulationObject TrackingRobust Privacy-Preserving Motion Detection and Object Tracking in Encrypted Streaming Video
Video privacy leakage is becoming an increasingly severe public problem, especially in cloud-based video surveillance systems. It leads to the new need for secure cloud-based video applications, where the video is encryp…
Motion DetectionMoving Object DetectionMultiple Object TrackingObject+5MambaTrack: Exploiting Dual-Enhancement for Night UAV Tracking
Night unmanned aerial vehicle (UAV) tracking is impeded by the challenges of poor illumination, with previous daylight-optimized methods demonstrating suboptimal performance in low-light conditions, limiting the utility …
GPUImage EnhancementMambaObject Tracking