Upload 7 files

Caption_it1.py

@@ -0,0 +1,71 @@
+# %%
+from keras.layers import *
+from keras.models import *
+from tensorflow.keras.models import Model
+from tensorflow.keras.applications import ResNet50
+from tensorflow.keras.preprocessing import image
+import numpy as np 
+from tensorflow.keras.applications.resnet50 import preprocess_input
+import pickle
+from tensorflow.keras.preprocessing.sequence import pad_sequences
+from tensorflow.keras.utils import to_categorical
+
+# %%
+model=load_model('model_weights/model_19.h5')
+
+# %%
+model_temp=ResNet50(weights='imagenet',input_shape=(224,224,3))
+
+# %%
+model_resnet=Model(model_temp.input,model_temp.layers[-2].output)
+
+# %%
+def preprocess_img(img):
+    img=image.load_img(img,target_size=(224,224))
+    img=image.img_to_array(img)
+    img=np.expand_dims(img,axis=0)
+
+    img=preprocess_input(img)
+
+    return img
+
+# %%
+def encode_image(img):
+    img=preprocess_img(img)
+    feature_vector=model_resnet.predict(img)
+    feature_vector=feature_vector.reshape(1,feature_vector.shape[1])
+    # print(feature_vector.shape)
+    return feature_vector
+
+# %%
+with open('storage/word_to_idx.pkl','rb') as w2i:
+    word_to_idx=pickle.load(w2i)
+with open('storage/idx_to_word.pkl','rb') as i2w:
+    idx_to_word=pickle.load(i2w)
+
+# %%
+def predict_caption(photo):
+    max_len=35
+    in_text = "startseq"
+    for i in range(max_len):
+        sequence = [word_to_idx[w] for w in in_text.split() if w in word_to_idx]
+        sequence = pad_sequences([sequence],maxlen=max_len,padding='post')
+        
+        ypred = model.predict([photo,sequence])
+        ypred = ypred.argmax() #WOrd with max prob always - Greedy Sampling
+        word = idx_to_word[ypred]
+        in_text += (' ' + word)
+        
+        if word == "endseq":
+            break
+    
+    final_caption = in_text.split()[1:-1]
+    final_caption = ' '.join(final_caption)
+    return final_caption
+
+# %%
+def caption_this_image(image):
+    enc=encode_image(image)
+    caption=predict_caption(enc)
+
+    return caption

app.py

@@ -0,0 +1,28 @@
+from flask import Flask, render_template, redirect, request
+
+import Caption_it1
+
+app=Flask(__name__)
+
+@app.route('/')
+def hello():
+    return render_template('index.html')
+
+@app.route('/',methods=['POST'])
+def marks():
+    if request.method=='POST':
+        f=request.files['userfile']
+        path='static/{}'.format(f.filename)
+        f.save(path)
+
+        caption=Caption_it1.caption_this_image(path)
+
+        result_dic={
+            'image':path,
+            'caption':caption
+        }
+
+    return render_template('index.html',your_result=result_dic)
+
+if __name__=='__main__':
+    app.run(debug=True)

encoded_test_features.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ac8012adb9e24405560b28392d70ea4b6dd34222a937b060352a7e838163c80
+size 8251187

encoded_train_features.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebb805645a60e973b859a64cbda1f634b9d797352dca6169305b2b75477f0d40
+size 49506244

requirements.txt

@@ -0,0 +1,4 @@
+Flask==2.3.3
+keras==2.10.0
+numpy==1.25.2
+tensorflow==2.10.0