Why prediction accuracy is higher with noisy images than denoised images?
2 views (last 30 days)
Show older comments
How to improve CNN accuracy with denoised images-please help
0 Comments
Answers (1)
Mahesh Taparia
on 15 Jun 2020
Hi
Adding noise in the training dataset act as a regularizer and it prevent overfitting. So the prediction accuracy is more.
To improve the accurcay, you can try with different architecture of CNN, use different optimizer, do data augmentation to increase the dataset, use different loss function for optimization, etc.
0 Comments
See Also
Categories
Find more on Recognition, Object Detection, and Semantic Segmentation in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!