In [22]:
import pandas as pd
import numpy as np
from collections import Counter
from matplotlib import pyplot as plt
import os
import cv2
%matplotlib inline
In [24]:
%%javascript
IPython.OutputArea.auto_scroll_threshold = 9999;
In [2]:
def load_filepaths():
imdir_ideology = 'ideology_image_dataset/'
ideology_files=os.listdir('ideology_image_dataset/')
ideology_files_path=[os.path.join(imdir_ideology,file) for file in ideology_files ]
return ideology_files_path
In [3]:
ideology_files_path=load_filepaths()
len(ideology_files_path)
Out[3]:
In [21]:
def showClustering(predicted_labels,label):
label_indexs= np.where(predicted_labels==label)[0]
print("CLUSTER--> ",label,"TOTAL IMAGES--> ",len(label_indexs))
if(len(label_indexs)>=500):
fig=plt.figure(figsize=(10, 400))
elif(len(label_indexs)>100 and len(label_indexs)<500):
fig=plt.figure(figsize=(10, 70))
elif(len(label_indexs)>=50 and len(label_indexs)<100):
fig=plt.figure(figsize=(10, 30))
elif(len(label_indexs)>=20 and len(label_indexs)<50):
fig=plt.figure(figsize=(10, 20))
elif(len(label_indexs)>=0 and len(label_indexs)<20):
fig=plt.figure(figsize=(10, 8))
for i,index in enumerate(label_indexs):
image = cv2.imread(ideology_files_path[index])
image= cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
columns = 4
rows = np.ceil(len(label_indexs)/float(columns))
fig.add_subplot(rows,columns, i+1)
plt.imshow(image)
plt.show()
IMAGE CLUSTERING RESULTS WITH GRIDSEARCH¶
VGG-16¶
The results of vgg16 in the dataset are shown below , we will not visualize the clustering reuslts because the rest resulted of few clusters
In [6]:
results_df=pd.read_csv('image-results/ideology_model_vgg_16.npy.csv')
results_df
Out[6]:
VGG-19¶
The results of vgg16 in the dataset are shown below , we will not visualize the clustering reuslts because the rest resulted of few clusters
In [12]:
results_df=pd.read_csv('image-results/ideology_model_vgg_19.npy.csv')
results_df
Out[12]:
Resnet-50¶
The results for resnet are below , we can see that agglomerative clustering with scipy resulted good number of clusters , lets visualize that
In [15]:
results_df=pd.read_csv('image-results/ideology_model_resnet50.npy.csv')
results_df
Out[15]:
In [25]:
predicted_Labels=np.load('image-results/ideology_model_resnet50.npy_agg-scipy_labels.npy')
unique_labels=set(predicted_Labels)
for label in unique_labels:
if(label!=-1):
showClustering(predicted_Labels,label)
