The paper trail for visual object tracking

What it does This repository is essentially a curated bibliography of visual object tracking research. It collects papers—from foundational surveys to recent conference work—alongside arXiv and official paper links, plus code repositories where available. The maintainer, a researcher at Zhejiang Normal University, also uses it to surface their own Transformer-tracking survey.

The interesting bit Rather than dumping titles, the list is grouped thematically and by venue, making it a decent pulse check on where the discipline is heading. You can watch the field’s priorities shift in real time: efficiency hacks like token pruning, neuromorphic spin-offs using spiking neural networks, and trackers that borrow from SAM or large vision-language models.

Key highlights

• Surveys and primers: includes major review papers on Siamese networks, discriminative filters, and deep learning for tracking.
• Efficiency-focused trackers: token-pruning frameworks (UTPTrack, ETCTrack) and lightweight transformers for UAV and edge scenarios.
• Neuromorphic and event-based: dedicated sections on spike-driven and event-camera trackers (SpikeTrack, SDTrack, EvoTrack).
• Multi-modal and language-guided: RGB-T trackers, retrieval-augmented language models (RAGTrack), and template-free vision-language approaches (MVLM).
• SAM ecosystem: links to SAM 2 derivatives like SAMURAI and segmentation-and-track hybrids.

Caveats

• Several listed papers have empty or missing code links, so this is a reading list, not a runnable toolkit.
• The repository itself contains no original implementation or benchmark code; it is purely a reference index.

Verdict Worth bookmarking if you are a tracking researcher or graduate student trying to map the literature. Skip it if you are hunting for a drop-in tracking library with pretrained weights.