AI Science Tracker
EN

Deep-learning image search helps identify three Nasca geoglyphs verified in field surveys

Yamagata University; IBM Japan · 1 Jun 2023 · humanities

Discovery

Deep-learning image search helps identify three Nasca geoglyphs verified in field surveys

Source check: Source checked. Not independent scientific confirmation.

Finding: A learned object detector trained on pictorial elements from 21 known geoglyphs guided human inspection of aerial-image candidates and subsequent on-site surveys. The 2023 primary report documents verified legs, fish and bird figures. These objects were already among the geoglyphs identified by 2022; this record dates the detailed 2023 report and does not claim they were first discovered that year. A fourth humanoid figure announced in 2019 is excluded from this record.

Technology: Deep-learning object detection on multiscale aerial imagery

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
A trained deep-learning detector learns pictorial components at multiple image scales and proposes aerial-image candidates for human review and field investigation, directly linking learned image search to the three recorded archaeological finds.
Test and result
Primary university Methods/results describe 307 element/multiscale training images from 21 known figures, human review of model detections and on-site verification of the legs, fish and bird geoglyphs.
Limitation / still unknown
The documented June 1,2023 report first publishes these three figures’ photographs and illustrations, but says they were already included among 358 geoglyphs identified by 2022. No absolute first-public-discovery priority is claimed. The source-reported 21-fold candidate-search speed comparison is not an independently replicated overall archaeological productivity gain. The earlier 2019 humanoid and separate 2024 catalog result of 303 geoglyphs are not counted here.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://www.sciencedirect.com/science/article/pii/S0305440323000559 · https://www.yamagata-u.ac.jp/en/files/4116/8559/7849/Nasca_Release_Paper_20230601_V10.pdf

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 21 Nov 2026. Review scope: Exact publisher title/DOI/abstract and full primary university two-page research Methods/results release. Original publisher full paper inaccessible; a third-party journal reproduction was only corroborative and is not treated as authenticated primary Methods..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse humanities discoveries.

Related catalog entries

Links use documented relationships where available, plus shared catalog context. A shared topic or model family does not establish a scientific dependency.