I have applied U-Net model for my images and it is giving good result. your videos are really very helpful. I have one request, please make a tutorial on ROC curve and confusion matrix with U-Net semantic segmentation. Please
Hi sir, thank u for ur videos, they are helping me a lot with my masters degree... everytime i need something, i search for ur videos and if isnt avaiable yet, u just realease in the week something i really need it... its like u are reading my mind and knowing my problems and gaps, and healing them... i really appreciate
You have to define the weights for the first layer if you want to use the pretrained weights from a model that originally accepts different channel images. In summary, you can use a pretrained model that got trained on RGB for grey scale images but for the first convolutional layer you need to load only weights from single channel. This can be a bit tricky. Therefore, I recommend converting your gray images to RGB, which is an easier approach.
My workdays (and weekends) are very busy so unfortunately I am unable to collaborate on any projects. I am sure you will gain enough knowledge from this channel to work on your projects. Also, you may want to look for faculty at your local Universities who may be able to mentor you.
I have applied U-Net model for my images and it is giving good result. your videos are really very helpful.
I have one request, please make a tutorial on ROC curve and confusion matrix with U-Net semantic segmentation. Please
Hi sir, thank u for ur videos, they are helping me a lot with my masters degree... everytime i need something, i search for ur videos and if isnt avaiable yet, u just realease in the week something i really need it... its like u are reading my mind and knowing my problems and gaps, and healing them... i really appreciate
Thank you very much.
same here !!! this channel helps so much for my master's degree
sir, pl can u share the video of stp by step procedure to create ourown dataset of satellite images for LULC using UNet.
What is the name of last video that you are referencing in this video?
Hello sir, is it possible to use pretrained models for grayscale images? Do I have to convert it into RGB before passing it into the model?
You have to define the weights for the first layer if you want to use the pretrained weights from a model that originally accepts different channel images. In summary, you can use a pretrained model that got trained on RGB for grey scale images but for the first convolutional layer you need to load only weights from single channel. This can be a bit tricky. Therefore, I recommend converting your gray images to RGB, which is an easier approach.
I always watch all your videos. You have been a great teacher. I will like to work with you on a project if that is okay.
My workdays (and weekends) are very busy so unfortunately I am unable to collaborate on any projects. I am sure you will gain enough knowledge from this channel to work on your projects. Also, you may want to look for faculty at your local Universities who may be able to mentor you.
next video?
These videos are truly invaluable. Thank you.
🤣🤣🤣🤣