A sleek Streamlit app for flower enthusiasts. With a choice between a custom-trained model and a fine-tuned ResNet50, users can effortlessly classify flowers—roses, daisies, dandelions, sunflowers, and tulips. Upload or pick random images to witness the magic of AI-driven classification.
The primary objective of FloraVision was to create a solo-developed, user-friendly Streamlit web application that harnesses the power of machine learning for flower classification. With a focus on providing an interactive experience, the project aimed to empower users to effortlessly categorize various flowers, including roses, daisies, dandelions, sunflowers, and tulips, utilizing both a custom-trained model and a fine-tuned ResNet50 model. By offering options to upload personal images or select randomly from a dataset, FloraVision sought to make flower identification engaging and accessible to all, highlighting the capabilities of AI in a visually appealing manner.
To see how I went from simple to complex custom architecture, check out the architecture directory
- Frontend: Streamlit
- Deep Learning: TensorFlow, VGG16, ResNet50
- Data Handling: Pandas, Numpy
- Data Visualization: Matplotlib, Seaborn