The MVP of Visual Product Recommendations is designed to be a game changer for the shopping experience. By leveraging advanced AI technologies, the app provides product recommendations based on user-provided images, focusing primarily on clothes and shoes.
The project employs a fine-tuned ImageNet model, specifically adapted to recognize a wide range of clothing and footwear. The dataset, collected from the Amazon website using a custom crawler, is enriched through computer vision algorithms to enhance prediction quality. The AI model operates as a REST service.
- Image-Based Recommendations: Users can take photos of clothes or shoes, and the app intelligently recommends similar products.
- AI-Driven Insights: Utilizes a fine-tuned ImageNet model, specifically optimized for high accuracy in clothing and footwear categorization.
- Data Augmentation: The dataset is augmented using computer vision techniques, to increase the quality of predictions.
- RESTful AI Model: The AI model is served as a REST service, allowing for easy integration and scalable architecture.
- User-Friendly Interface: The Android app provides a straightforward and intuitive interface for capturing photos and viewing recommendations.