Metadata-Version: 2.1
Name: analogcaption
Version: 0.1.0
Summary: A Package that uses AI to give users the caption of an image
Home-page: https://bitbucket.org/Lormenyoh/image-caption-generator/src/master/
Author: Analog Teams
Author-email: james@analogteams.com
License: The MIT License
Description: **Analog Image Caption Package**
        
        This is a Python package that helps give caption to your images, it is built on the Flickr_8k Dataset and trained with the Xception pre-trained model.
        
        
        
        ---
        
        ## How to try this package ?
        
        
        This package is very easy to use and has 2  fuctions which are:
        
        1. ***extract_features()*** this function simply take the image and extract the features from it.
        2. ***generate_desc()*** this function now generates a description for the Image.
        
        ---
        
        ## Demo in a flask app
        ```
        @app.route("/generateCaption", methods=["POST"])
        def generateCaption():
            image = request.files['image']
            img = image.read()
        
            # convert string of image data to uint8
            nparr = np.fromstring(img, np.uint8)
            # decode image
            img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
            img = cv2.resize(img, (224, 224))
           
        
            photo = extract_features(img)
            # generate description
            caption = generate_desc(model, tokenizer, photo, max_length)
        
           
            return render_template("results.html", image=image, caption=caption)
        ```
        ---
        
        ## Demo in a python script
        ```
        #import image and extract feature
        photo = extract_features(img_path)
        img = Image.open(img_path)
        description = generate_desc(model, tokenizer, photo, max_length)
        print("\n\n")
        print(description)
        plt.imshow(img)
        ```
        
        
        Hope this helps you understand the package and use it in a project. We are currently working on more cool features and will do an update when we're done.
Platform: UNKNOWN
Description-Content-Type: text/markdown
