Upload 3 files

app.py

@@ -0,0 +1,141 @@
+import streamlit as st
+from PIL import Image
+import numpy as np
+import nltk
+nltk.download('stopwords')
+nltk.download('punkt')
+import pandas as pd
+import random
+import easyocr
+import re
+from nltk.corpus import stopwords
+from nltk.tokenize import word_tokenize
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
+from transformers import AutoTokenizer, VisionEncoderDecoderModel, ViTFeatureExtractor
+@st.cache(allow_output_mutation=True)
+
+# Directory path to the saved model on Google Drive
+model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+
+# Load the feature extractor and tokenizer
+feature_extractor = ViTFeatureExtractor.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+tokenizer = AutoTokenizer.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+
+
+def generate_captions(image):
+    generated_caption = tokenizer.decode(model.generate(feature_extractor(image, return_tensors="pt").pixel_values.to("cpu"))[0])
+    sentence = generated_caption
+    text_to_remove = "<|endoftext|>"
+    generated_caption = sentence.replace(text_to_remove, "")
+    return generated_caption
+
+# use easyocr to extract text from the image
+def image_text(image):
+    reader = easyocr.Reader(['en'])
+    text = reader.readtext(np.array(image))
+    detected_text = " ".join([item[1] for item in text])
+
+    # Extract individual words, convert to lowercase, and add "#" symbol
+    detected_text= ['#' + entry[1].strip().lower().replace(" ", "_") for entry in text]
+    return detected_text
+
+# Load NLTK stopwords for filtering
+stop_words = set(stopwords.words('english'))
+
+# Add hashtags to keywords, which have been generated from image captioing
+def add_hashtags(keywords):
+    hashtags = []
+    
+    for keyword in keywords:
+        # Generate hashtag from the keyword (you can modify this part as per your requirements)
+        hashtag = '#' + keyword.lower()
+        
+        hashtags.append(hashtag)
+    
+    return hashtags
+
+def trending_hashtags():
+  # Read trending hashtags from a file separated by commas
+  with open("hashies.txt", "r") as file:
+      hashtags_string = file.read()
+
+  # Split the hashtags by commas and remove any leading/trailing spaces
+  trending_hashtags = [hashtag.strip() for hashtag in hashtags_string.split(',')]
+
+  # Create a DataFrame from the hashtags
+  df = pd.DataFrame(trending_hashtags, columns=["Hashtags"])
+
+  # Function to extract keywords from a given text
+  def extract_keywords(text):
+      tokens = word_tokenize(text)
+      keywords = [token.lower() for token in tokens if token.lower() not in stop_words]
+      return keywords
+
+  # Extract keywords from caption and trending hashtags
+  caption_keywords = extract_keywords(caption)
+  hashtag_keywords = [extract_keywords(hashtag) for hashtag in df["Hashtags"]]
+
+  # Function to calculate cosine similarity between two strings
+  def calculate_similarity(text1, text2):
+      tfidf_vectorizer = TfidfVectorizer()
+      tfidf_matrix = tfidf_vectorizer.fit_transform([text1, text2])
+      similarity_matrix = cosine_similarity(tfidf_matrix[0], tfidf_matrix[1])
+      return similarity_matrix[0][0]
+
+  # Calculate similarity between caption and each trending hashtag
+  similarities = [calculate_similarity(' '.join(caption_keywords), ' '.join(keywords)) for keywords in hashtag_keywords]
+
+  # Sort trending hashtags based on similarity in descending order
+  sorted_hashtags = [hashtag for _, hashtag in sorted(zip(similarities, df["Hashtags"]), reverse=True)]
+
+  # Select top k relevant hashtags (e.g., top 5) without duplicates
+  selected_hashtags = list(set(sorted_hashtags[:5]))
+
+  selected_hashtag = [word.strip("'") for word in selected_hashtags]
+
+  return selected_hashtag
+
+# create the Streamlit app
+def app():
+    st.title('Iamge from your Side, Trending Hashtags from our Site')
+
+    st.write('Upload an image to see what we have in store, Alwyas For You!')
+
+    # create file uploader
+    uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])
+
+    # check if file has been uploaded
+    if uploaded_file is not None:
+        # load the image
+        image = Image.open(uploaded_file)
+
+        # Image Captions
+        string = generate_captions(uploaded_file)
+        tokens = word_tokenize(string)
+        keywords = [token.lower() for token in tokens if token.lower() not in stop_words]
+        hashtags = add_hashtags(keywords)
+
+        # Text Captions from image
+        in_image_text = image_text(uploaded_file)
+
+        #Final Hashtags Generation
+        web_hashtags = trending_hashtags()
+
+        combined_hashtags = hashtags + in_image_text + web_hashtags
+
+        # Shuffle the list randomly
+        random.shuffle(combined_hashtags)
+
+        combined_hashtags = list(set(item for item in combined_hashtags[:15] if not re.search(r'\d$', item)))
+
+
+        # display the image
+        st.image(image, width=500, height=400)
+        st.write("Here it is THE CAPTIONS Just! for your Photo",string)
+        st.write("ohh! finally we have decided these are the best ",combined_hashtags)
+
+# run the app
+if __name__ == '__main__':
+    app()
+

hashies.txt

@@ -0,0 +1 @@
+'#instagood','#sand','#surfboard ','#love', '#photography', '#instagram', '#photooftheday', '#india', '#picoftheday', '#nature', '#instadaily', '#likeforlikes', '#follow', '#fashion', '#travel', '#followforfollowback', '#kerala', '#beautiful', '#style', '#art', '#travelphotography', '#followme', '#model', '#naturephotography', '#photo', '#smile', '#instalike', '#happy', '#bhfyp', '#like', '#life', '#stayhome', '#mumbai', '#cute', '#photoshoot', '#photographer', '#friends', '#me', '#instamood', '#food', '#likeforfollow', '#travelgram', '#lifestyle', '#keralagram', '#insta', '#trending', '#beauty', '#fun', '#like4like', '#igers', '#tiktok', '#like4likes', '#sunset', '#incredibleindia', '#throwback', '#girl', '#staysafe', '#keralatourism', '#likeforlike', '#bestoftheday', '#fashionblogger', '#kochi', '#quarantine', '#sky', '#portrait', '#wanderlust', '#keralagodsowncountry', '#tbt', '#likes', '#instafashion', '#instapic', '#followers', '#mobilephotography', '#music', '#fitness', '#mallu', '#delhi', '#swag', '#photographers_of_india', '#amazing', '#motivation', '#follow4follow', '#explore', '#godsowncountry', '#summer', '#followforfollow', '#modeling', '#loveyourself', '#indianphotography', '#indian', '#mountains', '#landscape', '#look', '#likesforlike', '#goodvibes', '#streetphotography', '#lockdown', '#travelblogger', '#selfie', '#malayalam', '#comment','#nyc', '#dogsofinstagram', '#california', '#cell', '#phone', '#newyork', '#losangeles', '#miami', '#family', '#florida', '#beach', '#foodie', '#adventure', '#explorepage', '#hiking', '#dog', '#foodporn', '#2020', '#makeup', '#socialdistancing', '#artist', '#newyorkcity', '#realestate', '#design', '#puppy', '#covid19', '#ootd', '#blacklivesmatter', '#roadtrip', '#interiordesign', '#outdoors', '#repost', '#blessed', '#colorado', '#vacation', '#texas', '#chicago', '#coronavirus', '#puppiesofinstagram', '#inspiration', '#quarantinelife', '#home', '#puppylove', '#flowers', '#dogstagram', '#nofilter', '#selfcare', '#selflove', '#sandiego', '#getoutside', '#fitnessmotivation', '#entrepreneur', '#smallbusiness', '#dogs', '#france', '#paris', '#sun', '#confinement', '#weekend', '#instamoment', '#sea', '#holidays', '#architecture', '#southoffrance', '#picture', '#vacances', '#marseille', '#igersfrance', '#bretagne', '#mood', '#lyon', '#frenchgirl', '#soleil', '#blackandwhite', '#cotedazur', '#provence', '#amour', '#frenchriviera', '#sunnyday', '#french', '#paysage', '#bordeaux', '#view', '#sport', '#europe', '#plage', '#naturelovers', '#instafood', '#instatravel', '#shooting', '#photographie', '#summervibes', '#nice', '#parisienne', '#trip', '#landscapephotography', '#outfit', '#blue', '#pictureoftheday', '#parisfrance', '#maldives', '#maldivesislands', '#ocean', '#paradise', '#holiday', '#indianocean', '#island','#islandlife', '#traveltheworld', '#maldivesresorts', '#beautifuldestinations', '#beachlife', '#traveling', '#travelling', '#honeymoon', '#maldiveslovers', '#luxurylifestyle', '#luxury', '#maldive', '#traveler', '#malediven', '#traveller', '#beautifulmaldives', '#luxurytravel', '#memories', '#maldivas', '#diving', '#scubadiving', '#tropical', '#relax', '#beachvibes', '#maldivesisland', '#underwaterphotography', '#sunnysideoflife', '#underwater', '#traveladdict', '#couplegoals', '#maldivesmania', '#resort', '#travelblog', '#maldivesinsider', '#clouds', '#sand'

requirements.txt

@@ -0,0 +1,11 @@
+streamlit==0.88.0
+Pillow==8.4.0
+numpy==1.21.4
+nltk==3.6.5
+pandas==1.3.4
+easyocr==1.4
+scikit-learn==0.24.2
+torch==1.9.0
+transformers==4.11.3
+altair<5
+click<8.0