GHCI : Innovative Solutions for the Specially-Abled people post COVID-19 life.
Link: https://ghci.azurewebsites.net/
Yolo v3 is an algorithm that uses deep convolutional neural networks to detect objects.
This project is written in Python 3.8.3 using Tensorflow (deep learning), NumPy (numerical computing),
OpenCV (computer vision) and seaborn (visualization) packages.
pip install -r requirements.txt
Let's download official weights pretrained on COCO dataset.
wget -P weights https://pjreddie.com/media/files/yolov3.weights
Now you can run the model using app.py script.
• Python
• OpenCV
• NumPy
• Math
Step 1: Detect people in the frame using YOLOv3 and depict it with bounding boxes.
Step 2:The pixel distance is calculated from the user’s device to the centre of the bounding box.
Arbitrary values have been considere for the project. Distance is recorded simultaneously
depending on the movement of the person.
As we know, Python has various applications and there are different libraries for different purposes. In our
further demonstration, we will be using the following libraries:
* Selenium: Selenium is a web testing library. It is used to automate browser activities.
* BeautifulSoup: Beautiful Soup is a Python package for parsing HTML and XML documents. It creates parse trees
that is helpful to extract the data easily.
Step 1: Find the URL that you want to scrape
Step 2: Inspecting the Page
Step 3: Find the data you want to extract
Step 4: Write the code
Step 5: Run the code and extract the data
Step 6: Store the data in a required format
The model uses the Web Speech API
Step 1:Select the “Start Recognition” button to start recording.
Step 2: The API then detects voice and converts into speech. If nothing is heard it gives out a message
for the user to start speaking again.
Step 3: The text is displayed on screen and can be saved as html or text files.
Step 4: Stop recording...
This project is written in Python 3.8.3 using Pyttsx3 (text-to-speech conversion library in Python.
Unlike alternative libraries, it works offline, and is compatible with both Python 2 and 3)
and PyPDF2 (Pure-Python library built as a PDF toolkit) libraries.
pip install pyttsx3
pip install PyPDF2
Step 1: Input a .pdf file from the user using html, css and js frontend.
Step 2: Store the uploaded pdf in the uploads folder using Flask.
Step 3: Open the file and read it using PyPDF2.PdfFileReader().
Step 4: Obtain the number of pages in the uploaded file.
Step 5: Initialize the speaker using pyttsx3.init().
Step 6: Extract text from each page using .extractText() and tell it out loud using speaker.say(text) and
speaker.runAndWait() commands.
This project is written in Python 3.8.3 using Pyttsx3 (text-to-speech conversion library in Python.
Unlike alternative libraries, it works offline, and is compatible with both Python 2 and 3) library.
pip install pyttsx3
Step 1: Input the text and male/female version choice from the user using html, css and js frontend.
Step 2: Store text and chosen option using flask.
Step 3: Initialize the speaker using pyttsx3.init().
Step 4: Set the voice rate and volume level using speaker.setProperty().
Step 5: Obtain the text and given choice using speaker.getProperty().
Step 6: Convert it to speech using speaker.say(text) and speaker.runAndWait() commands.
- PRASEEDHA PRAVEEN KALBHAVI
- MINI SHAIL CHHABRA