Search papers, labs, and topics across Lattice.
This paper introduces Tracking Identities with Dual-branch Elasticity (TIDE), a novel approach for tracking densely clustered and morphologically variable targets like schooling fish without relying on expensive appearance features. TIDE employs the Adaptive Geometric Correspondence IoU to maintain robust tracking through spatial and structural consistency, achieving a significant computational efficiency with a 38.7-fold reduction in FLOPs compared to state-of-the-art methods. Evaluations on the MFT-Edge benchmark show that TIDE's Lightweight L-branch achieves a competitive HOTA of 28.43, enabling real-time edge deployment while the Scalable S-branch reaches 29.98 HOTA, effectively balancing precision and efficiency.
TIDE achieves real-time tracking of schooling fish with a 38.7-fold reduction in computational overhead compared to leading methods, making it viable for edge deployment.
Tracking dense, homogeneous targets like schooling fish remains a major challenge for multiple object tracking due to extreme inter-individual homogeneity, severe physical clustering, and rapid non-rigid deformations. While heavy-backbone separated detection and embedding trackers like SU-T push accuracy boundaries using complex Re-Identification networks, their computational overhead prohibits edge deployment. Furthermore, these modules often fail when appearance features degrade under severe occlusions. To overcome this, we propose Tracking Identities with Dual-branch Elasticity (TIDE). Bypassing expensive appearance cues, TIDE utilizes the Adaptive Geometric Correspondence IoU, an association mechanism leveraging spatial and structural consistency to robustly handle complex morphological variations. Crucially, TIDE introduces system-level deployment elasticity, decoupling the algorithmic pipeline from strict hardware constraints. Evaluations on the MFT-Edge benchmark demonstrate that our Lightweight L-branch achieves a competitive HOTA of 28.43 using merely 20.47G FLOPs. This represents a 38.7-fold computational reduction compared to upper bounds like SU-T, directly facilitating real-time edge deployment. Concurrently, our Scalable S-branch establishes a 29.98 HOTA, successfully bridging the gap between high-precision cloud analysis and efficient edge tracking. The dataset and codes are released at https://vranlee.github.io/TIDE/.
