Načítá se...
Seamless lesion insertion for data augmentation in CAD training
The performance of a classifier is largely dependent on the size and representativeness of data used for its training. In circumstances where accumulation and/or labeling of training samples is difficult or expensive, such as medical applications, data augmentation can potentially be used to allevia...
Uloženo v:
| Vydáno v: | IEEE Trans Med Imaging |
|---|---|
| Hlavní autoři: | , , , |
| Médium: | Artigo |
| Jazyk: | Inglês |
| Vydáno: |
2016
|
| Témata: | |
| On-line přístup: | https://ncbi.nlm.nih.gov/pmc/articles/PMC5509514/ https://ncbi.nlm.nih.gov/pubmed/28113310 https://ncbi.nlm.nih.govhttp://dx.doi.org/10.1109/TMI.2016.2640180 |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo otaguje tento záznam!
|