Posts

Dual Privacy Protection for Distributed Fiber Sensing with Disaggregated Inference and Fine-tuning of Memory-Augmented Networks

We propose a memory-augmented model architecture with disaggregated computation infrastructure for fiber sensing event recognition. By leveraging geo-distributed computingresources in optical networks, this approach empowers end-users to customize models while ensuring dual privacy protection.