Update src/app.py

src/app.py

@@ -11,19 +11,16 @@ dataset = load_dataset("huggan/CelebA-faces")
 candidate_subset = dataset["train"].select(range(10)) # This is a small CBIR app! :D
 
 def index_dataset(dataset):
-    # This function might need to be split up, to reduce start-up time of app
-    # It could also use batches to increase speed
-    # If indexes are saved in files, this is all not really necessary
 
-    ## Color Embeddings
+    print("Color Embeddings")
     cd = ColorDescriptor((8, 12, 3))
-    dataset_with_embeddings = dataset.map(lambda row: {'color_embeddings': cd.describe(row["image"])}) # we assume that dataset has a column 'image'
+    dataset_with_embeddings = dataset.map(lambda row: {'color_embeddings': cd.describe(row["image"])})
 
-    ## CLIP Embeddings
+    print("CLIP Embeddings")
     clip_model = CLIPImageEncoder()
     dataset_with_embeddings = dataset_with_embeddings.map(clip_model.encode_images, batched=True, batch_size=16)
 
-    ## LBP Embeddings
+    print("LBP Embeddings")
     lbp_model = LBPImageEncoder(8,2)
     dataset_with_embeddings = dataset_with_embeddings.map(lambda row: {'lbp_embeddings': lbp_model.preprocess_img(row["image"])})
 
@@ -52,7 +49,6 @@ def check_index(ds):
 dataset_with_embeddings = check_index(candidate_subset)
 
 # Main function, to find similar images
-# TODO: allow different descriptor/embedding functions
 # TODO: implement different distance measures
 
 def get_neighbors(query_image, selected_descriptor, top_k=5):