Transferable Deployment of Semantic Edge Inference Systems via Unsupervised Domain Adaption

2025-04-16

AI Summary

To address high annotation costs and model retraining requirements in cross-environment deployment of semantic edge inference systems—caused by distribution shifts in sensor data and wireless channels—this paper proposes DASEIN, the first unsupervised domain adaptation (UDA) framework tailored to this setting. DASEIN introduces a two-stage joint alignment mechanism that simultaneously corrects source–target domain distribution discrepancies and channel distortion effects, operating entirely without target-domain labels. The method synergistically integrates UDA, knowledge distillation, and semantic feature encoding–decoding optimization to jointly satisfy lightweight IoT-side encoding and edge-side fused inference. Experiments demonstrate that under significant sensor distribution shifts, DASEIN improves inference accuracy in new environments by 7.09% (at matched SNR) and 21.33% (at 25 dB lower SNR) over the best baseline—first empirically validating the efficacy of co-adapting to both data and channel domain shifts in semantic edge inference transfer.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Application Domains: Internet of Things, Sensor Networks & Smart Cities
• Search and Optimization: Distributed Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

This paper investigates deploying semantic edge inference systems for performing a common image clarification task. In particular, each system consists of multiple Internet of Things (IoT) devices that first locally encode the sensing data into semantic features and then transmit them to an edge server for subsequent data fusion and task inference. The inference accuracy is determined by efficient training of the feature encoder/decoder using labeled data samples. Due to the difference in sensing data and communication channel distributions, deploying the system in a new environment may induce high costs in annotating data labels and re-training the encoder/decoder models. To achieve cost-effective transferable system deployment, we propose an efficient Domain Adaptation method for Semantic Edge INference systems (DASEIN) that can maintain high inference accuracy in a new environment without the need for labeled samples. Specifically, DASEIN exploits the task-relevant data correlation between different deployment scenarios by leveraging the techniques of unsupervised domain adaptation and knowledge distillation. It devises an efficient two-step adaptation procedure that sequentially aligns the data distributions and adapts to the channel variations. Numerical results show that, under a substantial change in sensing data distributions, the proposed DASEIN outperforms the best-performing benchmark method by 7.09% and 21.33% in inference accuracy when the new environment has similar or 25 dB lower channel signal to noise power ratios (SNRs), respectively. This verifies the effectiveness of the proposed method in adapting both data and channel distributions in practical transfer deployment applications.

Problem

• Deploying semantic edge systems across varied environments without labeled data
• Reducing costs of data annotation and model retraining in new settings
• Adapting to data and channel distribution changes for accurate inference

Innovation

• Unsupervised domain adaptation for semantic edge inference
• Two-step adaptation aligns data and channel distributions
• Knowledge distillation enhances task-relevant data correlation