Adaptive Semantic Token Selection for AI-native Goal-oriented Communications

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Devoto, Alessio, Petruzzi, Simone, Pomponi, Jary, Di Lorenzo, Paolo, Scardapane, Simone
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911866253475840
author Devoto, Alessio
Petruzzi, Simone
Pomponi, Jary
Di Lorenzo, Paolo
Scardapane, Simone
author_facet Devoto, Alessio
Petruzzi, Simone
Pomponi, Jary
Di Lorenzo, Paolo
Scardapane, Simone
contents In this paper, we propose a novel design for AI-native goal-oriented communications, exploiting transformer neural networks under dynamic inference constraints on bandwidth and computation. Transformers have become the standard architecture for pretraining large-scale vision and text models, and preliminary results have shown promising performance also in deep joint source-channel coding (JSCC). Here, we consider a dynamic model where communication happens over a channel with variable latency and bandwidth constraints. Leveraging recent works on conditional computation, we exploit the structure of the transformer blocks and the multihead attention operator to design a trainable semantic token selection mechanism that learns to select relevant tokens (e.g., image patches) from the input signal. This is done dynamically, on a per-input basis, with a rate that can be chosen as an additional input by the user. We show that our model improves over state-of-the-art token selection mechanisms, exhibiting high accuracy for a wide range of latency and bandwidth constraints, without the need for deploying multiple architectures tailored to each constraint. Last, but not least, the proposed token selection mechanism helps extract powerful semantics that are easy to understand and explain, paving the way for interpretable-by-design models for the next generation of AI-native communication systems.
format Preprint
id arxiv_https___arxiv_org_abs_2405_02330
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Adaptive Semantic Token Selection for AI-native Goal-oriented Communications
Devoto, Alessio
Petruzzi, Simone
Pomponi, Jary
Di Lorenzo, Paolo
Scardapane, Simone
Information Theory
Artificial Intelligence
Machine Learning
94A40
In this paper, we propose a novel design for AI-native goal-oriented communications, exploiting transformer neural networks under dynamic inference constraints on bandwidth and computation. Transformers have become the standard architecture for pretraining large-scale vision and text models, and preliminary results have shown promising performance also in deep joint source-channel coding (JSCC). Here, we consider a dynamic model where communication happens over a channel with variable latency and bandwidth constraints. Leveraging recent works on conditional computation, we exploit the structure of the transformer blocks and the multihead attention operator to design a trainable semantic token selection mechanism that learns to select relevant tokens (e.g., image patches) from the input signal. This is done dynamically, on a per-input basis, with a rate that can be chosen as an additional input by the user. We show that our model improves over state-of-the-art token selection mechanisms, exhibiting high accuracy for a wide range of latency and bandwidth constraints, without the need for deploying multiple architectures tailored to each constraint. Last, but not least, the proposed token selection mechanism helps extract powerful semantics that are easy to understand and explain, paving the way for interpretable-by-design models for the next generation of AI-native communication systems.
title Adaptive Semantic Token Selection for AI-native Goal-oriented Communications
topic Information Theory
Artificial Intelligence
Machine Learning
94A40
url https://arxiv.org/abs/2405.02330