We are pleased to announce that our latest research has been published in Scientific Reports (Nature Portfolio, Open Access).
The study, titled “Enhancing identity stability in sperm motility analysis via density-adaptive tracklet stitching with YOLOv12n and BoostTrack++,” proposes an automated sperm motility analysis framework that integrates the YOLOv12n detector, the BoostTrack++ multi-object tracker and a novel Density-Adaptive Tracklet Stitching (DATS) post-tracking module that reconnects fragmented tracklets according to local sperm density.
Evaluated on our Hi-LabSpermTracking dataset in three scenarios — detector comparison across six models, tracker comparison under temporally separated train/test conditions, and generalization to videos from unseen individuals — the framework achieves consistently strong HOTA, MOTA and IDF1 scores, and DATS further reduces identity fragmentation without altering the tracking architecture. The result is a practical solution for trajectory-based motility analysis from conventional bright-field microscopy videos.
Authors:
İmran Gül (Yıldız Technical University)
Hamza Osman İlhan (Yıldız Technical University)
Hakkı Uzun (Recep Tayyip Erdoğan University)
Merve Hüner Yiğit (Recep Tayyip Erdoğan University)
Nizamettin Aydın (Istanbul Technical University)
Görkem Serbes (Yıldız Technical University)
