媒体计算实践作业：图像——文本跨模态搜索

Cottontail DB is a column store vector database aimed at multimedia retrieval. It allows for classical boolean as well as vector-space retrieval (nearest neighbour search) used in similarity search using a unified data and query model.

ADAMpro is a database system to store and retrieve multimedia data. It provides Boolean retrieval and similarity search and makes use of a various index structures for efficient retrieval.

Long-Tail Hashing (Learning to hash from more realistic long-tail datasets), SIGIR 2021, CCF A类.

vitrivr NG is a web-based user interface for searching and browsing mixed multimedia collections. It uses cineast as a backend

Adversarial queries for blocking Content-based Image Retrieval (CBIR)

vitrivr-VR is a virtual reality interface for the vitrivr multimedia retrieval system.

Various utilities to deal with metadata and content provided by the Berlin State Library/Staatsbibliothek zu Berlin

vitrivr's next-generation retrieval engine. It is capable of extracting and retrieving a wider range of multimedia objects such as audio, video, images or 3d models.

Paper summary of Most read paper in Advanced Topics in Multimedia Analysis and Indexing course in National Taiwan University

A simple attention deep learning model to answer questions about a given video with the most relevant video intervals as answers.

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops

In-memory index-structures for efficient retrieval of multimedia documents based on input queries.

ImaginAItion is a small app for image retrieval that I developed as the final project for the Multimedia Data Management course at the University of Bologna.

An implementation of unsupervised hypergraph-based manifold ranking for image analysis.

Inter-Modal Retrieval

ObjectLens: Intelligent 2D & 3D object detection and retrieval system combining YOLOv8 real-time detection with multi-modal feature extraction (form, texture, color) and FAISS-powered similarity search. Built with React/Vite frontend and FastAPI backend for scalable multimedia mining and 3D-aware image indexing.

A Java-based system that identifies video frames by hashing and indexing them in JSON files. Given a query frame from an .rgb video, the system retrieves the matching video and timestamp, then plays the result using a JavaFX media player. Features multithreading for optimized search

🌟 Accelerate SQLite database operations on Apple Silicon with GPU power for faster analytics and efficient data management.