{
  "id": 9689992,
  "title": "TrackEverything: Long Horizon Dense Tracking via De-Duplicating 3D Scene Representations",
  "url": "https://urgent.news/2026/09/24/trackeverything-long-horizon-dense-tracking-via-de-duplicating-3d",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-24T17:48:20.000Z",
  "source": {
    "name": "arXiv cs.AI",
    "slug": "arxiv-cs-ai",
    "url": "https://arxiv.org/abs/2609.30222v1"
  },
  "original_language": "en",
  "account": null,
  "summary": "Existing point tracking models face a fundamental tradeoff: they can either track a sparse set of query points over long horizons, or track all points across only short clips. We introduce TrackEverything, a 3D point tracker that breaks this trade-off by representing videos as persistent 3D scene tracks in world coordinates. Grounded in the insight that videos are 2D projections of an underlying…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}