About
Our Mission
At SmartBookshelf.io, we're focused on simplifying the process of cataloging and organizing personal libraries. Whether you're managing your books for personal use, preparing for sale, or keeping track of textbooks for college, our advanced machine learning technologies can help. Our platform allows you to create and manage personal collections, easily add or remove books, and organize your documents. Our LLM algorithm then works to match each extracted text with a book in our database, making your book management more efficient and streamlined.
Our Technology
SmartBookshelf.io uses a combination of Optical Character Recognition (OCR) and machine learning to detect and extract text from book spines. Here's a breakdown of our technology stack:
- Frontend: We use React, Next.js, Tailwind CSS, and DaisyUI to build a responsive and user-friendly interface.
- Backend: Our backend is powered by Python and Flask, providing fast and reliable endpoints for image processing and text extraction.
- OCR: We utilize Google Cloud Vision API and Tesseract OCR to accurately detect and extract text from images of book spines.
- Machine Learning: Our system uses YOLO (You Only Look Once) for real-time object detection, ensuring that we accurately identify and crop book spines from uploaded images.
Other Features in Progress
- AI Detector: An upcoming feature to detect AI-generated text within book descriptions and analyses.
- Plagiarism Detector: A feature to ensure the originality of the text by detecting plagiarism in essays and other writings.
- Essay Writer: Generate essays based on selected books with customizable formats, styles, and keyword focuses.
- GPU support: Integrate a GPU to speed up the process for all users.
- Async updates: Improve the speed of the algorithm by fully utilizing caching and preloading of predictions.
Frequently Asked Questions
What is SmartBookshelf.io?
SmartBookshelf.io is a web application that allows you to upload images of your bookshelf and extract text from the book spines using OCR (Optical Character Recognition) technology.
How do I use SmartBookshelf.io?
Simply upload an image of your bookshelf, and our application will detect the books and extract the titles for you. An algorithm is in development that will automatically detect the books and extract additional information for you.
What technologies are used in SmartBookshelf.io?
SmartBookshelf.io uses modern web technologies such as React, Next.js, Tailwind CSS, and DaisyUI for the front end. The backend is powered by Python, Flask, and various OCR libraries.
How can I ensure the best results when uploading images?
Ensure your photo is well-lit, in focus, and that book titles are clearly visible. It is best to separate books of the same size or color as they may blend into one otherwise.
What if the book titles are not detected accurately?
You can manually adjust any incorrect book titles after the analysis. We are working on improving the accuracy of our system.
Crop Feature
The crop feature allows you to select specific areas of your uploaded images to focus on. This is particularly useful if the image contains multiple books and you want to refine the detection process. Simply upload an image, use the cropping tool to select the desired areas, and then click "Add Crop" to save your selection. You can add multiple crops if needed.
Disclaimer
SmartBookshelf.io is a project in development, and while we strive to provide accurate and efficient results, there may be inaccuracies in text detection and extraction. For images with poor cropping, bad lighting or tons of books, the results returned by the system may not be accurate. However, it should give you a rough estimate on at least the number of books regardless. We appreciate your understanding and welcome any feedback to help us improve the system.
Known Bugs
We are currently aware of an issue where the color of the cropped book spines may not match the original image. This is a known bug and we are working on a fix. We appreciate your patience and understanding as we work to resolve this issue.
Get in Touch
If you have any questions, feedback, or suggestions, feel free to reach out using the feedback form. You can also reach me at rob@smartbookshelf.io.