HIGHLIGHTS
- who: George Zhou from the Weill Cornell Medicine, New York, NY, USA have published the paper: On the analysis of data augmentation methods for spectral imaged based heart sound classification using convolutional neural networks, in the Journal: (JOURNAL)
- what: With these constraints in mind, the authors evaluate common data augmentation techniques at the audio level, including pitch shifting and time stretching/compressing and noise injection, and at the image level, including horizontal flips, vertical flips, hue/brightness transformations, principal component analysis (PCA) color augmentation, random color filters, and time/frequency masking, for classification of heart . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.