mean Corruption Error (mCE): 30.8Number of params: 99M
Papers with Code · CC BY-SA 4.0
Domain Generalization
The idea of Domain Generalization is to learn from one or multiple training domains, to extract a domain-agnostic model which can be applied to an unseen domain
<span class="description-source">Source: Diagram Image Retrieval using Sketch-Based Deep Learning and Transfer Learning </span>
논문
mean Corruption Error (mCE): 28.2Number of params: 1100M
mean Corruption Error (mCE): 38.74Number of params: 87M
mean Corruption Error (mCE): 35.8Top 1 Accuracy: 73.6Number of params: 77M
mean Corruption Error (mCE): 36.80Number of params: 87M
mean Corruption Error (mCE): 38.8Number of params: 350M
mIoU: 58.01
mIoU: 67.75
mean Corruption Error (mCE): 31.4Number of params: 632M
mean Corruption Error (mCE): 42.1Top 1 Accuracy: 69.2Number of params: 77M
mIoU: 43.5
mean Corruption Error (mCE): 39.0
mean Corruption Error (mCE): 38.4Number of params: 296M
mIoU: 55.31
mean Corruption Error (mCE): 33.8Number of params: 632M
mean Corruption Error (mCE): 42.5
mIoU: 46.99
mIoU: 65.6
mIoU: 66.4