Browse fragrances with images and descriptions
a tiny vision language model
Explore news topics through interactive visuals
Generate dynamic torus knots with random colors and lighting
Image captioning, image-text matching and visual Q&A.
View and submit results to the Visual Riddles Leaderboard
Display a loading spinner and prepare space
Visualize 3D dynamics with Gaussian Splats
Ask questions about images to get answers
Rerun viewer with Gradio
Generate descriptions and answers by combining text and images
Answer questions based on images and text
Answer questions about images in natural language
Danielle is an advanced Visual QA (Question Answering) tool designed to assist users in exploring and understanding fragrances through images and descriptions. It leverages AI technology to provide detailed insights and recommendations based on visual and textual inputs, making it a unique solution for fragrance discovery.
• Image-based browsing: Danielle allows users to upload images of fragrances or browse through a database to find matching scents.
• Description analysis: The tool provides detailed descriptions of fragrances, including notes, accords, and usage occasions.
• AI-powered recommendations: Based on user preferences or image inputs, Danielle suggests similar or complementary fragrances.
• Real-time Q&A: Users can ask specific questions about fragrances, and Danielle will provide accurate answers based on its extensive database.
What types of images does Danielle support?
Danielle supports high-quality images of fragrance bottles, labels, or packaging. Clear images with visible text or branding yield the best results.
Can Danielle analyze perfume without an image?
Yes, Danielle can work with text descriptions or questions about fragrances. Simply type your query to get started.
Is Danielle available in multiple languages?
Currently, Danielle is available in English, with plans to expand to other languages in the future.