Skip to content

Latest commit

 

History

History
33 lines (21 loc) · 1.64 KB

File metadata and controls

33 lines (21 loc) · 1.64 KB

Visualization-of-activations-features-of-CNN-layers-

In this repository I will show you how we can visualize the features extracted from the intermediate convolutional layers form different models in keras models.

PRE REQUISITES:

  1. Tensorflow
  2. Keras
  3. Python

Please make sure that you download the weights of the pre-trained models from this repository and paste these weights files in C:/Users/yourname/ .keras/models/ directory.

INPUT IMAGE

Input Image

Visualization of the features/activations from the convolutional layers in the RESNET 50 CNN network

Visualization of the Feature Maps Extracted from block 1 in the RESNET_50 model Convolutional Block 1 features/activations

Visualization of the Feature Maps Extracted from block 2 in the RESNET_50 model Convolutional Block 2 features/activations

Visualization of the Feature Maps Extracted from block 3 in the RESNET_50 model Convolutional Block 3 features/activations

Visualization of the Feature Maps Extracted from block 4 in the RESNET_50 model Convolutional Block 4 features/activations

Visualization of the Feature Maps Extracted from block 5 in the RESNET_50 model Convolutional Block 5 features/activations

Observations to be made: