image-embeddings

computer vision and sports

Train Models Contrastively in Pytorch

Kernel Fisher Discriminant Analysis implementation following https://arxiv.org/abs/1906.09436

A lightweight Text-to-Image Retrieval model [Web App]

LightweightEmbeddings is a fast, free, and unlimited API service for multilingual embeddings and reranking, with support for both text and images and guaranteed uptime.

Code implementation for our ICPR, 2020 paper titled "Improving Word Recognition using Multiple Hypotheses and Deep Embeddings"

Code implementation for our DAS, 2020 paper titled "Fused Text Recogniser and Deep Embeddings Improve Word Recognition and Retrieval"

Jina CLIP v2 - Multimodal embedding model for text and images with Matryoshka representations (64-1024 dimensions). Deployed on Replicate with optimized GPU inference.

SmartGallery is a desktop app that organizes and searches large photo collections using AI. It generates captions, tags, and CLIP embeddings, supports natural language search, live folder updates, and fast thumbnail viewing, turning scattered images into a structured, searchable personal gallery.

Raven – The Embedder is a standalone Python module from the Retraven project that powers text and image embedding pipelines. It integrates with Qdrant for vector storage, MinIO for object storage, and RabbitMQ for asynchronous task orchestration—making it ideal for embedding-driven AI applications.

Assess Data Quality Before Annotation or Labelled Data Quality after Annotation (Txt files/Yolo Format). Visualise the patterns covered by each class/activity.

🏀 Discover real-time sports insights with SportsFeed AI, an app that uses AI to answer your sports questions using free APIs and open-source models.

Reverse Image Search for Shopify Products

Multimodal extractors for video, image, audio, text & PDF — turn any file into searchable vector embeddings (SigLIP, Gemini, E5, CLAP, ArcFace).

Calculate image and document images on edge. Use these embeddings for on-edge use cases and flow them to our system for other uses.

Reverse image search using EfficientNet-b2 embeddings and FAISS for fast similarity matching. Web UI with Streamlit.

Multimodal Knowledge Retrieval System with Optional Memory (MKRS)

🖼️ Generate high-quality multimodal embeddings for text and images with Jina CLIP v2, offering flexible dimensions and optimized performance for diverse applications.