提出球面角度间隔机制，提升图像离散表示的多样性与覆盖度。
ArcVQ-VAE: A Spherical Vector Quantization Framework with ArcCosine Additive Margin
- 引入球面角度间隔先验，约束码本向量分布并增强角距离。
- 在图像重建与生成任务中实现更优的表征多样性和样本质量。
- 适合需要高效利用码本、追求高质量生成的视觉表征学习场景。
向量量化变分自编码器（VQ-VAE）已成为图像建模中学习离散表征的基础框架。然而，传统VQ-VAE需用有限码本向量对整张图像进行编码，容量限制导致难以捕捉丰富多样的表征。本文提出弧余弦加边距向量量化框架（ArcVQ-VAE），在标准VQ-VAE码本上引入球面角度间隔先验（SAMP）。该先验包含球形范数正则化（约束所有码本向量位于随时间变化的欧氏球内）与弧余弦加边距损失（促进潜在向量间的角分离）。该设计提升了受限空间内潜在表示的判别性与均匀性，从而改善有效潜空间覆盖率并提升码本利用率。在标准图像重建与生成任务上的实验表明，ArcVQ-VAE在重建精度、表征多样性与样本质量方面均达到与基线模型相当的性能。代码已公开于：https://github.com/goals4292/ArcVQ-VAE
原文摘要 · Abstract (English)
Vector Quantized Variational Autoencoder (VQ-VAE) has become a fundamental framework for learning discrete representations in image modeling. However, VQ-VAE models must tokenize entire images using a finite set of codebook vectors, and this capacity limitation restricts their ability to capture rich and diverse representations. In this paper, we propose ArcCosine Additive Margin VQ-VAE (ArcVQ-VAE), a novel vector quantization framework that introduces a spherical angular-margin prior (SAMP) for the codebook of a conventional VQ-VAE. The proposed SAMP consists of Ball-Bounded Norm Regularization, which constrains all codebook vectors within a time-dependent Euclidean ball, and ArcCosine Additive Margin Loss, which encourages greater angular separability among latent vectors. This formulation promotes more discriminative and uniformly dispersed latent representations within the constrained space, thereby improving effective latent-space coverage and leading to improved codebook utilization. Experimental results on standard image reconstruction and generation tasks show that ArcVQ-VAE achieves competitive performance against baseline models in terms of reconstruction accuracy, representation diversity, and sample quality. The code is available at: https://github.com/goals4292/ArcVQ-VAE
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。