Enabling Federated Inference via Unsupervised Consensus Embedding

2026-05-07

AI Summary

This work addresses the challenge of enabling collaborative inference among multiple models without sharing raw data, model parameters, or a common encoder. The authors propose the Consensus Embedding Federated Inference (CE-FI) framework, which employs an unsupervised consensus embedding layer to align heterogeneous intermediate representations into a unified embedding space and a collaborative output layer to produce joint predictions—requiring only unlabeled data for training. CE-FI is the first method to achieve privacy-preserving cross-model collaborative inference without assuming shared model architectures or labeled data, thereby overcoming the strong sharing assumptions inherent in conventional approaches. Experiments demonstrate that CE-FI significantly outperforms single-model inference under non-IID settings on CIFAR-10/100, matches the performance of strong-sharing baselines, and exhibits robust generalization across multimodal tasks including image, text, and time-series data.

Abstract

Cooperative inference across independently deployed machine learning models is increasingly desirable in distributed environments, as there is a growing need to leverage multiple models while keeping their data and model parameters private. However, existing cooperative frameworks typically rely on sharing input data, model parameters, or a common encoder, which limits their applicability in privacy-sensitive or cross-organizational settings. To address this challenge, we propose Consensus Embedding-based Federated Inference (CE-FI), a framework that enables pretrained models to cooperate at inference time without sharing model parameters or raw inputs and without assuming a common encoder. CE-FI introduces two components: a Consensus Embedding (CE) layer that maps heterogeneous intermediate representations into a common embedding space, and a Cooperative Output (CO) layer that produces predictions from these embeddings. Both layers are trained using shared unlabeled data only, so the cooperative stage does not require additional labeled data. Experiments on image classification benchmarks -- CIFAR-10 and CIFAR-100 -- under diverse non-IID conditions show that CE-FI consistently outperforms solo inference and performs comparably to conventional methods that require stronger sharing assumptions. Additional evaluations on text and time-series tasks indicate applicability beyond image classification, although performance depends on the ensemble strategy. Further analysis identifies representation alignment as the primary bottleneck.

Problem

• Federated Inference
• Privacy-Preserving Cooperation
• Model Collaboration
• Heterogeneous Models
• Unsupervised Embedding

Innovation

• Federated Inference
• Consensus Embedding
• Privacy-Preserving Collaboration
• Unsupervised Alignment
• Heterogeneous Models