MVTrack: Ultrafast Appearance-Free Moving Object Tracking from Compressed Bitstreams

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Iñaki Erregue, Kamal Nasrollahi, Sergio Escalera
• Venue: arXiv: cs.CV
• Published: 2026-08-11
• Collected: 2026-08-12
• URL: http://arxiv.org/abs/2608.10790v1
• PDF: https://arxiv.org/pdf/2608.10790v1
• DOI: UNKNOWN
• Registry group: academic_apis_and_preprints
• Evidence role: canonical_academic_index
• Language: en
• Country / region: UNKNOWN

Plain-Language Brief

The researchers present MVTrack, a fast moving object tracker that works directly on H.264 bitstreams without needing to reconstruct pixels. It outperforms YOLO26n in terms of parameters, FLOPs, and CPU latency when tested on VIRAT dataset.

What the Document Says

• The researchers introduced MVTrack, a fast moving object tracker.
• MVTrack operates on H.264 bitstreams without pixel reconstruction.
• It combines MVDet for motion vector field detection and MVLink for minimal kinematic association.
• Compared to YOLO26n, MVTrack uses fewer parameters, requires fewer FLOPs, and reduces CPU latency by 8.6 times on VIRAT dataset.

Key Concepts Mentioned

• MVTrack
• H.264 bitstreams
• moving object tracking
• MVDet
• MVLink
• YOLO26n
• VIRAT
• parameters
• FLOPs
• CPU latency

Claims Made by the Document

• MVTrack is an ultrafast tracker for moving objects.
• MVTrack outperforms YOLO26n on VIRAT dataset.

Evidence Used

• Results demonstrated by the researchers

Methods / Approach

• Operating on compressed video data
• Using MVDet and MVLink

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Deploying modern video trackers at scale is bottlenecked by the computational cost of RGB-based object detectors. To this end, we present MVTrack, an ultrafast tracker for moving objects that operates directly on H.264 bitstreams. MVTrack combines MVDet, a lightweight detector for motion vector fields, with MVLink, a minimalist kinematic association module. On VIRAT, MVTrack outperforms YOLO26n while using 60$\times$ fewer parameters, requiring 40$\times$ fewer FLOPs, and reducing CPU latency by 8.6$\times$. These results demonstrate that compressed video data alone can enable accurate and scalable surveillance tracking, thereby bypassing the need for pixel reconstruction.