---
title: "Dreams Are Cheaper Than Flights: Curated World-Model Data Improves Aerial Detection"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.41030v1"
permalink: "https://feed.astralor.com/items/itm_3c563bdf0a2591e98a98d685"
published: "2026-10-29T22:17:32.000Z"
score: 53
channel: "tech"
tags: ["多模态","Dreams Are Cheaper","Than Flights","Curated World-Model Data","Improves Aerial Detection","Robot","Video"]
---

# Dreams Are Cheaper Than Flights: Curated World-Model Data Improves Aerial Detection

## At a glance

Robot perception is increasingly data driven, yet field data is slow, costly, and sometimes disruptive to collect. Video world models offer a cheaper source: seeded with real frames, they extend a recorded scene into ne…

## Key points

- Robot perception is increasingly data driven, yet field data is slow, costly, and sometimes disruptive to collect.
- Video world models offer a cheaper source: seeded with real frames, they extend a recorded scene into new footage with the deployment's sensor, altitude, and environment.
- We ask whether such synthetic data matches the utility of real field data and whether detectors trained with it perform in the field.

## Full text

Robot perception is increasingly data driven, yet field data is slow, costly, and sometimes disruptive to collect. Video world models offer a cheaper source: seeded with real frames, they extend a recorded scene into new footage with the deployment's sensor, altitude, and environment. We ask whether such synthetic data matches the utility of real field data and whether detectors trained with it perform in the field. Seeding a 7B-parameter world model with wildlife drone video, we more than double the training set with curated frames. On held-out videos from the same drones, mAP@50--95 rises 4.3\% when these frames are added to real data and 10.3\% when they replace it, and up to 33.5\% on an urban benchmark. Although an expert rejects an estimated 87.5\% of generated clips, the data costs about \$1,100 of GPU time, against more than \$3,200 of field labor alone for a less useful real training set. Synthetic data thus shifts data costs from fieldwork to compute.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_3c563bdf0a2591e98a98d685