Advances in Efficient and Private Nearest Neighbor Search

The field of nearest neighbor search is witnessing significant developments, with a focus on improving efficiency and privacy. Recent research has explored innovative approaches to accelerate high-dimensional nearest neighbor search, including dynamic query preference and GPU-accelerated graph indexing. Additionally, there is a growing emphasis on balancing privacy and efficiency in music information retrieval and other applications, with techniques such as additive homomorphic encryption and privacy-preserving index design gaining traction. Noteworthy papers in this area include:

  • Accelerating High-Dimensional Nearest Neighbor Search with Dynamic Query Preference, which proposes a novel dual-index query framework for efficient search.
  • Balancing Privacy and Efficiency: Music Information Retrieval via Additive Homomorphic Encryption, which presents a practical approach to privacy-preserving music information retrieval.
  • Scalable Graph Indexing using GPUs for Approximate Nearest Neighbor Search, which introduces a fast library for graph indexing accelerated by GPUs.
  • Privacy-Preserving Approximate Nearest Neighbor Search on High-Dimensional Data, which introduces a novel solution for privacy-preserving k-ANNS on vectors.

Sources

Accelerating High-Dimensional Nearest Neighbor Search with Dynamic Query Preference

Balancing Privacy and Efficiency: Music Information Retrieval via Additive Homomorphic Encryption

Scalable Graph Indexing using GPUs for Approximate Nearest Neighbor Search

Succinct Oblivious Tensor Evaluation and Applications: Adaptively-Secure Laconic Function Evaluation and Trapdoor Hashing for All Circuits

Privacy-Preserving Approximate Nearest Neighbor Search on High-Dimensional Data

Built with on top of