<div style="text-align:center"> Photo by <a style="background-color:black;color:white;text-decoration:none;padding:4px 6px;font-family:-apple-system, BlinkMacSystemFont, "San Francisco", "Helvetica Neue", Helvetica, Ubuntu, Roboto, Noto, "Segoe UI", Arial, sans-serif;font-size:12px;font-weight:bold;line-height:1.2;display:inline-block;border-radius:3px" href="https://unsplash.com/photos/vJwzxQK7FaE" target="_blank" rel="noopener noreferrer" title="Download free do whatever you want high-resolution photos from Craig McKay "><span style="display:inline-block;padding:2px 3px"><svg xmlns="http://www.w3.org/2000/svg" style="height:12px;width:auto;position:relative;vertical-align:middle;top:-2px;fill:white" viewBox="0 0 32 32"><title>unsplash-logo</title><path d="M10 9V0h12v9H10zm12 5h10v18H0V14h10v9h12v-9z"></path></svg></span><span style="display:inline-block;padding:2px 3px">Craig McKay</span></a> </div>
Keypoint :
- Noisy Labels
Using methods :
- Relabeled dataset
- Noisy of health class is highest which contains 0,1,2,3 class.
- 0 ~ 1 class
- 2 ~ 3 class
- CutMix, FixMix, SnapMix
- Sharpness-Aware Minimization
- Bi-Tempered Logistic Loss
- Seed, Multi-Scale