Just an addition, at 5:44, the reason why the output image will not "perfectly" be same as the input image is that the code space has fewer dimensions compared to the input space. The reason for choosing the code space to have fewer dimensions than the input space is to avoid duplicating the input which is a simple way to achieve the task that we are discussing.
While training an Autoencoder to clean noisy images in the output do we have the clean / non - noisy images with us or are they not present. In case we have the clean images with us beforehand how does the Loss function look in that case ? Do we take in account the Reconstruction Loss or is the output of the AutoEncoder directly compared to the clean image that we have
Please Explain that part about Reconstruction of Images (Domain Adaptation). In that case how is the Auto Encoder trained ? Considering the example - If a Light Microscope Image is given as input that will be reconstructed right ? Then how does the Image from a Electron Microscope is generated ? How does the Loss Function look like
Just an addition, at 5:44, the reason why the output image will not "perfectly" be same as the input image is that the code space has fewer dimensions compared to the input space. The reason for choosing the code space to have fewer dimensions than the input space is to avoid duplicating the input which is a simple way to achieve the task that we are discussing.
sir you are literally a life saver , thank you so much ...
very nice material thanks,im waiting denoising with autoencoder :)
While training an Autoencoder to clean noisy images in the output do we have the clean / non - noisy images with us or are they not present.
In case we have the clean images with us beforehand how does the Loss function look in that case ?
Do we take in account the Reconstruction Loss or is the output of the AutoEncoder directly compared to the clean image that we have
i cant believe an indian guy can speak English like this. that's awesome
Please Explain that part about Reconstruction of Images (Domain Adaptation). In that case how is the Auto Encoder trained ?
Considering the example -
If a Light Microscope Image is given as input that will be reconstructed right ? Then how does the Image from a Electron Microscope is generated ?
How does the Loss Function look like
I have a question? Can we use it for grayscale IHC images
Nice. Thanks
GREAT LECTURES
Isn’t instance segmentation with unet example of domain adaptation? It maps image into segments using CNN. 🤔
Thank you!
You're welcome!
its amazing video
thank you Sir
Thank you.
You're welcome!
Thank you!