update multiple object

clip_component.py

@@ -4,9 +4,9 @@ import os
 from PIL import Image
 import clip
 
+similarity_threshold = 22.00
 def get_token_from_clip(image):
-
-    text_inputs = ["Bacon", "Bread", "Fruit", "Beans and Rice", "fries", "Lasagna"]
+    text_inputs = ["apple", "banana", "lemon", "orange", "cereal", "salad", "chicken", "juice", "milk", "bread"]
     text_tokens = clip.tokenize(text_inputs)
 
     device = "cpu"
@@ -27,15 +27,19 @@ def get_token_from_clip(image):
     with torch.no_grad():
         similarity = text_features.cpu().numpy() @ image_feature.cpu().numpy().T
 
-    best_similarity = 0
-    best_text_input = ""
+    results = []
+    detect_food = ""
 
     for i in range(similarity.shape[0]):
         similarity_num = (100.0 * similarity[i][0])
-        if similarity_num > best_similarity:
-            best_similarity = similarity_num
-            best_text_input = text_inputs[i]
-
-    # Print the caption for the image
-    print("Best caption for the image: ", best_text_input)
-    return best_text_input
+        text_input = text_inputs[i]
+        results.append({"text_input": text_input, "similarity": similarity_num})
+        if similarity_num >= similarity_threshold:
+            detect_food += " " + text_input + " ."
+        # print(similarity_num)
+    detect_food_list = detect_food[1:]
+    results.sort(key=lambda x: x["similarity"], reverse=True)
+    # Print the caption for each text input along with their similarity scores
+    for result in results:
+        print(f"Text input: {result['text_input']}, Similarity: {result['similarity']:.2f}")
+    return detect_food_list