Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Zero-Shot Knowledge Base Resizing for Rate-Adaptive Digital Semantic Communication

Shumin Yao, Hui Du, Lifeng Xie, Yaping Sun, Hao Chen, Nan Ma, Xiaodong Xu

TL;DR

This work tackles granular rate adaptation in VQ-VAE-based semantic communication by introducing zero-shot KB resizing. It leverages hyperbolic geometry to reveal a semantic hierarchy within a large, pre-trained KB, builds a master semantic tree via a minimum spanning tree, and deterministically prunes leaves to obtain any target KB size without retraining. The resized KB is mapped back to Euclidean space for compatibility, enabling on-the-fly rate control with minimal performance loss. Empirical results on ImageNet show near-parity with dedicated KBs trained from scratch and superior robustness at very low rates, while dramatically reducing training and storage costs. Overall, the approach provides a practical, scalable path toward truly rate-adaptive semantic communication systems.

Abstract

Digital semantic communication systems, which often leverage the Vector Quantized Variational Autoencoder (VQ-VAE) framework, are pivotal for future wireless networks. In a VQ-VAE-based semantic communication system, the transmission rate is directly governed by the size of a discrete codebook known as knowledge base (KB). However, the KB size is a fixed hyperparameter, meaning that adapting the rate requires training and storing a separate model for each desired size -- a practice that is too computationally and storage-prohibitive to achieve truly granular rate control. To address this, we introduce a principled, zero-shot KB resizing method that enables on-the-fly rate adaptation without any retraining. Our approach establishes a global importance ranking for all vectors within a single, large parent KB by uncovering its inherent semantic hierarchy. This is achieved via a three-step framework: 1) embedding KB vectors into hyperbolic space to reveal their hierarchical relationships; 2) constructing a master semantic tree using a minimum spanning tree algorithm; 3) enabling instant resizing by iteratively pruning the least important leaf nodes. Extensive simulations demonstrate that our method achieves reconstruction quality nearly identical to that of dedicated KBs trained from scratch, while demanding only a fraction of the computational budget. Moreover, our approach exhibits superior robustness at very low rates, where conventional KBs suffer from catastrophic failure. Our work resolves a fundamental limitation of VQ-VAE-based semantic communication systems, offering a practical and efficient path toward flexible and rate-adaptive semantic communication.

Zero-Shot Knowledge Base Resizing for Rate-Adaptive Digital Semantic Communication

TL;DR

This work tackles granular rate adaptation in VQ-VAE-based semantic communication by introducing zero-shot KB resizing. It leverages hyperbolic geometry to reveal a semantic hierarchy within a large, pre-trained KB, builds a master semantic tree via a minimum spanning tree, and deterministically prunes leaves to obtain any target KB size without retraining. The resized KB is mapped back to Euclidean space for compatibility, enabling on-the-fly rate control with minimal performance loss. Empirical results on ImageNet show near-parity with dedicated KBs trained from scratch and superior robustness at very low rates, while dramatically reducing training and storage costs. Overall, the approach provides a practical, scalable path toward truly rate-adaptive semantic communication systems.

Abstract

Digital semantic communication systems, which often leverage the Vector Quantized Variational Autoencoder (VQ-VAE) framework, are pivotal for future wireless networks. In a VQ-VAE-based semantic communication system, the transmission rate is directly governed by the size of a discrete codebook known as knowledge base (KB). However, the KB size is a fixed hyperparameter, meaning that adapting the rate requires training and storing a separate model for each desired size -- a practice that is too computationally and storage-prohibitive to achieve truly granular rate control. To address this, we introduce a principled, zero-shot KB resizing method that enables on-the-fly rate adaptation without any retraining. Our approach establishes a global importance ranking for all vectors within a single, large parent KB by uncovering its inherent semantic hierarchy. This is achieved via a three-step framework: 1) embedding KB vectors into hyperbolic space to reveal their hierarchical relationships; 2) constructing a master semantic tree using a minimum spanning tree algorithm; 3) enabling instant resizing by iteratively pruning the least important leaf nodes. Extensive simulations demonstrate that our method achieves reconstruction quality nearly identical to that of dedicated KBs trained from scratch, while demanding only a fraction of the computational budget. Moreover, our approach exhibits superior robustness at very low rates, where conventional KBs suffer from catastrophic failure. Our work resolves a fundamental limitation of VQ-VAE-based semantic communication systems, offering a practical and efficient path toward flexible and rate-adaptive semantic communication.
Paper Structure (8 sections, 7 equations, 4 figures)

This paper contains 8 sections, 7 equations, 4 figures.

Table of Contents

  1. Introduction
  2. System Model
  3. Zero-Shot Knowledge Base Resizing
  4. Hyperbolic embedding for revealing semantic granularity
  5. Hierarchical pruning for zero-shot resizing
  6. Euclidean projection for system compatibility
  7. Evaluation
  8. Conclusion

Figures (4)

  • Figure 1: Considered VQ-VAE-based digital semantic communication framework.
  • Figure 2: Procedures of the proposed zero-shot knowledge base resizing
  • Figure 3: SSIM versus KB size $K$: Our zero-shot resizing method achieves nearly identical performance to the baseline of training dedicated KBs from scratch, without requiring retraining for each size.
  • Figure 4: Qualitative comparison of reconstructed images under different KB sizes.