New answers tagged

0 votes

MS Azure Computer Vision Service Performing Worse on More Data

Custom Vision can behave this way if the increased data is confusing the underlying model. It can be due to wrongly labeled data. You can do a manual quality check of the newer data being added by ...
user avatar
0 votes

UpConvolution feature extraction

Upconvolutional layers are used to expand the spatial size of the input. So they do not "generate" features. It's just resizing itself. It does generates some numbers but these numbers are ...
user avatar
  • 157
1 vote

Loss decreases, but Validation Loss just fluctuates

You have 250 images as training set, and you are using a model with millions of parameters... I'm pretty sure that your model is just memorizing the training set, aka you are overfitting. At this ...
user avatar
2 votes
Accepted

Loss decreases, but Validation Loss just fluctuates

It looks like your model is overfitting: it's learning from the training dataset, but this learning doesn't apply to the test dataset. You can try to reduce the complexity of the model by simplifying ...
user avatar
  • 489
1 vote
Accepted

High level-Low Level features in U-NET

To answer your question let's first go through how CNN works. When we give a CNN an input image, it sees an array of numbers that correspond to the pixel intensities of the input image. The intensity ...
user avatar
  • 1,271
1 vote

Advice on vision ML classifier pipeline

Do you want to do multi-modal classification? If yes, the basic principle is to have one model for stats data, one model for image data, and one model that decides which one is more informative. ...
user avatar

Top 50 recent answers are included