Upload 3 files

utils/data_users.py

@@ -1,5 +1,7 @@
 import streamlit as st 
+import streamlit.components.v1 as components
 import os  
+# import cv2
 
 ROOT_FIG_DIR = f'{os.getcwd()}/figures/'
 def get_product_dev_page_layout():
@@ -68,25 +70,22 @@ def get_product_dev_page_layout():
     with row4_2:
         st.subheader('Exploratory Data Stats')
         with st.expander('Training Data Info'):
-            col1, col2 = st.columns((1,1))
-            st.write("**Raw Data Details**")
-            with col1:
-                st.caption('Number of Images per Category')
-                st.image(f'{ROOT_FIG_DIR}/train_raw_category_dist.png')
-            with col2:
-                st.caption('Aspect ratio')
-                st.image(f'{ROOT_FIG_DIR}/train_raw_aspectratio_dist.png')
-            st.caption('Image Samples')
-            st.image(f'{ROOT_FIG_DIR}/train_samples.png')
+            HtmlFile = open(f'{ROOT_FIG_DIR}/train_set_report.html', 'r', encoding='utf-8')
+            source_code = HtmlFile.read() 
+            components.html(source_code,scrolling=True, height=500)
 
         with st.expander('Test Data Info'):
-            st.write("**Raw Data Details**")
-            st.caption('Number of Images per Category')
-            st.image(f'{ROOT_FIG_DIR}/test_category_dist.png')
-            st.caption('Aspect ratio')
-            st.image(f'{ROOT_FIG_DIR}/test_aspectratio_dist.png')
-            st.caption('Image Samples')
-            st.image(f'{ROOT_FIG_DIR}/train_samples.png')
+            HtmlFile = open(f'{ROOT_FIG_DIR}/test_set_report.html', 'r', encoding='utf-8')
+            source_code = HtmlFile.read() 
+            components.html(source_code,scrolling=True, height=500)
+        
+        with st.expander('Sample Visualization for each Category'):
+            clss = st.selectbox('Select a category(class)', ["CNV","DME", "NORMAL", "DRUSEN"])
+            img_path = f'{ROOT_FIG_DIR}/{clss}_samples.png'
+            st.image(img_path)
+            # HtmlFile = open(f'{ROOT_FIG_DIR}/test_set_report.html', 'r', encoding='utf-8')
+            # source_code = HtmlFile.read() 
+            # components.html(source_code,scrolling=True, height=500)
 
     with row4_3:
         # st.write("**Post Processesd Data Details**")
@@ -107,17 +106,9 @@ def get_product_dev_page_layout():
         st.code(code, language='python')
 
         with st.expander('Training Data Info after Representative Sampling'):
-            # new_title = '<h5 style="color:Black;">Training Set Info after Representative Sampling:</h5>'
-            # st.markdown(new_title, unsafe_allow_html=True)
-            col1, col2 = st.columns((1,1))
-            with col1: 
-                st.caption('Number of Images per Category')
-                st.image(f'{ROOT_FIG_DIR}/train_filter_category_dist.png')
-            with col2:
-                st.caption('Aspect ratio')
-                st.image(f'{ROOT_FIG_DIR}/train_filter_aspectratio_dist.png')
-            st.caption('Image Samples')
-            st.image(f'{ROOT_FIG_DIR}/train_samples.png')
+            HtmlFile = open(f'{ROOT_FIG_DIR}/filtered_set_report.html', 'r', encoding='utf-8')
+            source_code = HtmlFile.read() 
+            components.html(source_code,scrolling=True, height=500)
         st.write("Model Input Size Resizing: 180x180x3")
         # st.caption('Post Processing Steps:')
         # code = '''def representative_sampling():

utils/decisions_users.py

@@ -86,11 +86,47 @@ def annoy_matching(annoy_f,query_item, query_index, n=10):
 def get_img(fn ,thumbnail=False):
     img = Image.open(fn)
     if thumbnail:
-        img.thumbnail((100,100))
+        img.thumbnail((150,150))
     return img
 
+def open_gray(fn):
+    img = cv2.cvtColor(cv2.imread(fn), cv2.COLOR_BGR2GRAY)
+    img = cv2.cvtColor(img, cv2.COLOR_GRAY2RGB)
+    img = cv2.resize(img,(224,224))
+    return img
 
 def plot_n_similar(seed_id,similar_ids, test_path,n=10, scale=5):
+    # img_list = []
+    # title_list = ["SEED ID:{0} <br> Label:{1}".format(seed_id,os.path.basename(test_labels[seed_id]))]
+    # # for indx,row in cnv_cls_df.iterrows():
+    # img_list.append(open_gray(test_path.replace("F:/","E:/")))
+    # for i in range(len(similar_ids)):
+    #     print("PATH:",similar_ids[i].replace("F:/","E:/"))
+    #     img_list.append(open_gray(similar_ids[i].replace("F:/","E:/")))
+    #     # ax[i+1].imshow(get_img(similar_ids[i].replace("F:/","E:/")),cmap='gray')
+    #     title = "ID:{0} <br> Distance: {1:.3f} <br> Label:{2}".format(i,0.1223,os.path.basename(similar_ids[i])[:-4])
+    #     title_list.append(title)  
+
+
+    # fig = px.imshow(np.array(img_list), facet_col=0, binary_string=True,facet_row_spacing=0.002,facet_col_spacing=0.002)
+    # # Set facet titles
+    # for i, sigma in enumerate(title_list):
+    #     fig.layout.annotations[i]['text'] = sigma
+    #     fig.layout.annotations[i]['yshift'] = -40
+    #     # fig.layout.tex
+    # fig.update_layout(
+    #     margin=dict(
+    #         l=10,
+    #         r=10,
+    #         b=10,
+    #         t=40,
+    #         pad=1
+    #     ),
+    # )
+    # fig.update_xaxes(showticklabels=False)
+    # fig.update_yaxes(showticklabels=False)
+    # fig.show()
+
     f,ax = plt.subplots(1,n+1,figsize=((n+1)*scale,scale))
     # print(os.path.basename(test_labels[seed_id])[:-4])
     title = "SEED ID:{0}\nLabel:{1}".format(seed_id,os.path.basename(test_labels[seed_id]))
@@ -98,17 +134,19 @@ def plot_n_similar(seed_id,similar_ids, test_path,n=10, scale=5):
     ax[0].imshow(get_img(test_path.replace("F:/","E:/")),cmap='gray')
     ax[0].set_title(title,fontsize=12)
     for i in range(len(similar_ids)):
+        # print("PATH:", similar_ids[i])
         ax[i+1].imshow(get_img(similar_ids[i].replace("F:/","E:/")),cmap='gray')
         title = "ID:{0}\nDistance: {1:.3f}\nLabel:{2}".format(i,0.1223,os.path.basename(similar_ids[i])[:-4])  
         ax[i+1].set_title(title,fontsize=10)
     f.suptitle("Images similar to seed_id {0}".format(seed_id),fontsize=18)
-    plt.subplots_adjust(top=0.5)
+    plt.subplots_adjust(top=0.4)
+    plt.tight_layout()
     return f
 
 def load_image(filename,change_url=True):
     # if change_url:
-    print(filename)
-    print(os.path.exists(filename))
+    # print(filename)
+    # print(os.path.exists(filename))
     img = cv2.imread(filename)
     return img
 
@@ -178,7 +216,8 @@ def run_instance_exp(img_path, img_path_list,prob_list,grad_vis_path_list):
     closest_idxs = annoy_matching(annoy_tree,query_item, seed_id, 10)
     closest_fns = [train_ids[close_i] for close_i in closest_idxs]
     st.subheader('Top-10 Similar Samples from Gallery Set')
-    st.pyplot(plot_n_similar(seed_id,closest_fns, img_path,n=10, scale=4))
+    st.plotly_chart(plot_n_similar(seed_id,closest_fns, img_path,n=10, scale=4), use_container_width=True)
+    # st.pyplot(plot_n_similar(seed_id,closest_fns, img_path,n=10, scale=4))
 
 
 
@@ -197,7 +236,8 @@ def run_instance_exp_keras_model(img_path, new_model, feature_extractor_model):
     display_image = load_image(img_path)
     # fig = px.imshow(display_image)
     # left_column.plotly_chart(fig, use_container_width=True)
-    left_column.image(cv2.resize(display_image, (180,180)),caption = "Selected Input")
+    left_column.image(display_image,caption = "Selected Input")
+    # left_column.image(cv2.resize(display_image, (180,180)),caption = "Selected Input")
    
 
     roi_img, probs = get_predictions_and_roi(img_path, new_model)
@@ -208,13 +248,14 @@ def run_instance_exp_keras_model(img_path, new_model, feature_extractor_model):
     prob_cls =np.asarray(probs)[0]
     # print(prob_cls)
     tmp_df = pd.DataFrame.from_dict({'class':LABELS,'probability':prob_cls})
-    fig = plt.figure(figsize=(5, 4))
+    fig = plt.figure(figsize=(8, 9))
     sns.barplot(x='probability', y='class', data=tmp_df)
     middle_column.pyplot(fig)
     # middle_column.write("Probabilities")
 
     # grad img 
-    right_column.image(roi_img, caption = "Decision ROI")
+    # right_column.image(roi_img, caption = "Decision ROI")
+    right_column.image(cv2.resize(roi_img, display_image.shape[:2]),caption = "GradCAM ROI")
 
     # seed_id = 900
     seed_id = test_id_list.index(org_img_path)
@@ -228,7 +269,8 @@ def run_instance_exp_keras_model(img_path, new_model, feature_extractor_model):
                 for each_fn in closest_fns]
     # print(closest_fns)
     st.subheader('Top-10 Similar Samples from Gallery Set')
-    st.pyplot(plot_n_similar(seed_id,closest_fns_tmp, img_path,n=10,scale=3))
+    # st.plotly_chart(plot_n_similar(seed_id,closest_fns, img_path,n=10, scale=4), use_container_width=True)
+    st.pyplot(plot_n_similar(seed_id,closest_fns_tmp, img_path,n=10,scale=4))
 
 def main():
 

utils/eval_users.py

@@ -36,13 +36,22 @@ def get_product_dev_page_layout():
         #     # st.caption('Test Set Results:')
         st.subheader('Test Set Confusion Matrix')
         st.image(f'{ROOT_FIG_DIR}/test_confmat_20210404.png')
-        st.subheader('Prediction Samples')
-        st.caption('Correctly Classified sample predictions:')
-        st.image(f'{ROOT_FIG_DIR}/pred_stats.png')
+        # st.subheader('Prediction Samples')
+        # st.caption('Correctly Classified sample predictions:')
+        # st.image(f'{ROOT_FIG_DIR}/pred_stats.png')
 
-        st.caption('Miss Classified sample predictions:')
-        st.image(f'{ROOT_FIG_DIR}/pred_stats.png')
+        # st.caption('Miss Classified sample predictions:')
+        # st.image(f'{ROOT_FIG_DIR}/pred_stats.png')
 
-        st.subheader("Class-wise Prediction Distributions")
-        st.image(f'{ROOT_FIG_DIR}/training_prob_stats.png')
+        # st.subheader("Class-wise Prediction Distributions")
+        # st.image(f'{ROOT_FIG_DIR}/training_prob_stats.png')
+    
+    with row6_3:
+        st.write("Weencountered classimbalance issue and here is the miclassified samples...")
+
+        st.caption('Miss Classified CNV Samples:')
+        st.image(f'{ROOT_FIG_DIR}/cnv_missclass.png')
+
+        st.caption('Miss Classified NORMAL Samples:')
+        st.image(f'{ROOT_FIG_DIR}/normal_missclass.png')