paper-with-me

Papers

Residual Quantization with Implicit Neural Codebooks

2024-01-26 · Iris A. M. Huijben, Matthijs Douze, Matthew Muckley, Ruud J. G. van Sloun, Jakob Verbeek

Vector quantization is a fundamental operation for data compression and vector search. To obtain high accuracy, multi-codebook methods represent each vector using codewords across several codebooks. Residual quantization (RQ) is one such method, which iteratively quantizes the error of the previous step. While the error distribution is dependent on previously-selected codewords, this dependency is not accounted for in conventional RQ as it uses a fixed codebook per quantization step. In this paper, we propose QINCo, a neural RQ variant that constructs specialized codebooks per step that depend on the approximation of the vector from previous steps. Experiments show that QINCo outperforms state-of-the-art methods by a large margin on several datasets and code sizes. For example, QINCo achieves better nearest-neighbor search accuracy using 12-byte codes than the state-of-the-art UNQ using 16 bytes on the BigANN1M and Deep1M datasets.

📄 PDF Abstract BibTeX arXiv:2401.14732

Code (1)

facebookresearch/qinco 공식 구현 pytorch

Tasks

Data CompressionQuantization

Similar Papers 제목 키워드 기반

Qinco2: Vector Compression and Search with Improved Implicit Neural Codebooks

2025-01-06 · Théophane Vallaeys, Matthew Muckley, Jakob Verbeek, Matthijs Douze

Vector quantization is a fundamental technique for compression and large-scale nearest neighbor search. For high-accuracy operating points, multi-codebook quantization associates data vectors with one element from each o…

DecoderQuantization

Semantic Codebooks as Effective Priors for Neural Speech Compression

2025-12-25 · Liuyang Bai, Weiyi Lu, Li Guo arxiv

Speech codecs are traditionally optimized for waveform fidelity, allocating bits to preserve acoustic detail even when much of it can be inferred from linguistic structure. This leads to inefficient compression and subop…

RQ-MoE: Residual Quantization via Mixture of Experts for Efficient Input-Dependent Vector Compression

2026-05-14 · Zhengjia Zhong, Shuyan Ke, Zaizhou Lin, Jiaqi Song 외 arxiv

Vector quantization is a fundamental tool for compressing high-dimensional embeddings, yet existing multi-codebook methods rely on static codebooks that limit expressiveness under heterogeneous data geometry. While recen…

Regularized Residual Quantization: a multi-layer sparse dictionary learning approach

2017-05-01 · Sohrab Ferdowsi, Slava Voloshynovskiy, Dimche Kostadinov

The Residual Quantization (RQ) framework is revisited where the quantization distortion is being successively reduced in multi-layers. Inspired by the reverse-water-filling paradigm in rate-distortion theory, an efficien…

Dictionary LearningQuantizationSuper-Resolution

Variable Bitrate Residual Vector Quantization for Audio Coding

2024-10-08 · Yunkee Chae, Woosung Choi, Yuhta Takida, Junghyun Koo 외

Recent state-of-the-art neural audio compression models have progressively adopted residual vector quantization (RVQ). Despite this success, these models employ a fixed number of codebooks per frame, which can be subopti…

Audio CompressionQuantization