HDRL-MoE is a hierarchical DRL framework with MoE that decouples slow inference decisions from fast UAV trajectory control in a constrained POMDP to maximize inference accuracy under mission constraints.
Early-exit deep neural network - a comprehensive survey,
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
verdicts
UNVERDICTED 2representative citing papers
A P4-SDN architecture uses a split early-exit CNN in the data plane and GRU in the control plane for real-time DDoS detection with reduced latency.
citing papers explorer
-
UAV-Assisted Cooperative Edge Inference for Low-Altitude Economy via MoE-based Hierarchical Deep Reinforcement Learning
HDRL-MoE is a hierarchical DRL framework with MoE that decouples slow inference decisions from fast UAV trajectory control in a constrained POMDP to maximize inference accuracy under mission constraints.
-
Collaborative P4-SDN DDoS Detection and Mitigation with Early-Exit Neural Networks
A P4-SDN architecture uses a split early-exit CNN in the data plane and GRU in the control plane for real-time DDoS detection with reduced latency.