Inspiration
Our motivation comes from using technology to improve the lives of those who are blind or visually impaired. Our objective is to develop a user-friendly solution that supports inclusivity, equality, and independence for blind people by leveraging AI, machine learning, computer vision, and other pertinent technologies. Our dedication lies in utilizing technology to improve accessibility and enable people with visual impairments to confidently and independently explore their surroundings, obtain information, and carry out daily duties.
What it does
The idea behind our blind person assistant project is to use technology to improve the lives of those who are blind or visually impaired. To give a full and encompassing toolset, our solution integrates advanced functions including object detection, text messaging, phone call capabilities, facial recognition for detecting familiar people, and label reading. Our objective is to develop an all-in-one solution that enables blind people to communicate, navigate their surroundings, and obtain information with ease by fusing cutting-edge technologies like AI, machine learning, and computer vision. This will ultimately foster independence and inclusivity.
How we built it
Object Detection: We utilized computer vision techniques and pre-trained machine learning models to detect objects in the environment. We used popular libraries such as OpenCV and TensorFlow to implement object detection capabilities. Text Messaging and Phone Call: We integrated with APIs or libraries that allow sending text messages and making phone calls programmatically. For example, we used Twilio for SMS and phone call functionalities, which enabled our solution to send text messages and make phone calls for communication purposes. Facial Recognition: We employed facial recognition algorithms to detect and recognize familiar faces. We used libraries such as OpenCV and dlib for facial recognition tasks, and trained our own facial recognition model using labeled data. Label Reading: We utilized optical character recognition (OCR) techniques to read and interpret labels on objects. We used libraries such as Tesseract, which is a popular OCR library in Python, to extract text from images and recognize labels. Backend and Frontend Development: We built the backend of our solution using Python and Flask, a web development framework, to handle server-side logic, API integrations, and data processing. For the frontend, we used HTML, CSS, and JavaScript to create a user-friendly interface for interacting with our solution. Testing and Iteration: We conducted thorough testing and iterative development to ensure the reliability, accuracy, and usability of our solution. We gathered feedback from visually impaired individuals to make improvements and refine our solution based on their needs and preferences.
Challenges we ran into
The integration and connecting of all the components was one of the difficulties we faced when developing our project for a blind person assistance. This required the smooth and user-friendly integration of object detection, face recognition, OCR, text messaging, phone calls, and other features. It was important to make sure that every part functioned as a whole and gave visually impaired people a consistent experience, which called for careful coordination and integration of numerous technologies and APIs.
Accomplishments that we're proud of
We are proud of creating a user-friendly and straightforward system in our blind person assistant project. This includes designing an intuitive interface and integrating all the functionalities in a seamless manner, making it easy for visually impaired users to interact with the system and utilize its features effectively. Our focus on accessibility and inclusivity has resulted in a solution that is user-friendly and accommodating for individuals with visual impairments, making it a significant accomplishment for our project.
What we learned
We learned a lot about combining several APIs and integrating them into a coherent system while working on our blind person assistance project. Additionally, we learnt how to optimize the system for quick runtimes in order to guarantee seamless and effective operation. The project improved our skills in application development, machine learning, computer vision, and human-computer interface by giving us practical experience working with a variety of technologies and APIs. We also developed a greater comprehension of the difficulties encountered by people with vision impairments and the significance of developing solutions that are accessible. All things considered, our team learned a great deal from the project, which helped us increase our technical expertise and acquire useful abilities for creating inclusive and empowering solutions.
What's next for AssistiveViewTech
The family recognition component of Ozmo, our blind person assistance project, will be further improved and refined in the next months. Our goal is to improve the facial recognition system's accuracy and dependability so that it can more precisely identify several family members. This could entail giving the machine learning models more training, optimizing the algorithms, and incorporating user feedback to promote ongoing development. In order to provide customers with more seamless interaction and customization possibilities, we also intend to investigate possible integration with other functionalities, such as speech recognition. Our objective is to develop Ozmo into a complete and trustworthy helper that can offer important support and aid to those who are blind or visually impaired in their daily lives.
Log in or sign up for Devpost to join the conversation.