retrieval

A high performance Vector Database leveraging parallel graph computing

Semantic retrieval for AI agents: high-recall memory and code search that cuts coding-agent token usage by nearly 50%

An Unsupervised Distance Learning Framework for Multimedia Retrieval

A image retrieval program implemented by VLAD algorithm (paper: https://ieeexplore.ieee.org/document/5540039). Can train you own image set and build your own image retrieval system.

Bumped Ribbon Retrieval and Approximate Membership Query

source code for the paper "Hard-Aware-Deeply-Cascaed-Embedding"

A Bayesian Nested-Sampling Retrieval Code

Skeleton-Based Canonical Forms for Non-Rigid 3D Shape Retrieval

SOMHunter core library.

High-performance ANN vector search engine in modern C++ implementing HNSW, IVF, PQ, SQ8, IVF-SQ, and IVF-PQ with SIFT1M benchmarking.

SCIAMACHY 2d satellite trace gas retrieval

A tool for interaction log collection in limited SOMHunter evnironment.

Living KV-cache: recall across 800k tokens on an 8GB GPU. Python PoC + llama.cpp port (no kernel fork) with one-command install. Core idea = ArkVale (NeurIPS 2024); see README.

Query expansion tools based on the co-ocurrence of word usage

RCP — the Retrieval Context Protocol. An open JSON-RPC standard so any RAG engine can expose embed/rerank/retrieve/graph and any client can consume it. The retrieval companion to MCP and ACP.

some scripts

High-performance, thread-safe embedding library using llama.cpp.

advanced concepts of data, storage, organization, and retrieval. Topics include multiple-linked lists, balanced trees, graphs, abstract data types, classes and methods, object-oriented programming, searching and sorting.

Embedded property graph database built for speed. Vector search and full-text search built in. Implements Cypher.

vsag is a vector indexing library used for similarity search.