ScholarGate
Asistent
Machine learningDeep Learning, Object Detection, Meta-Learning

Detekcija objekata sa malo primera (FSOD)

Detekcija objekata sa malo primera (Few-Shot Object Detection, FSOD) je pristup meta-učenja koji omogućava detekciju novih klasa objekata iz samo nekoliko anotiranih primera. Za razliku od standardne detekcije objekata koja zahteva stotine označenih instanci po klasi, FSOD uči da brzo prilagodi modele detekcije novim kategorijama objekata korišćenjem znanja iz osnovnih kategorija.

Otvorite u MethodMindUskoroVideoUskoroDownload slides

Pročitajte celu metodu

Samo za članove

Prijavite se besplatnim nalogom da biste pročitali ovaj odeljak.

Prijavite se

Method map

The neighbourhood of related methods — select a node to explore.

Detekcija objekata sa malo primera (FSOD)
DETR (Detection Transfor…SimCLRSwin Transformer

Izvori

  1. Wang, X., Huang, T. E., Darrell, T., Gonzalez, J. E., & Yu, F. (2020). Few-shot object detection with attention-RPN and multi-relation detector. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 9050-9059). link

Kako citirati ovu stranicu

ScholarGate. (2026, June 3). Few-Shot Object Detection with Contrastive Learning. ScholarGate. https://scholargate.app/sr/deep-learning/few-shot-object-detection

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

Compare side by side

Citirana u

ScholarGateFew-Shot Object Detection (Few-Shot Object Detection with Contrastive Learning). Preuzeto 2026-06-15 sa https://scholargate.app/sr/deep-learning/few-shot-object-detection · Skup podataka: https://doi.org/10.5281/zenodo.20539026