On Accelerating Edge AI: Optimizing Resource-Constrained Environments

This survey presents a comprehensive overview of strategies for accelerating deep learning models including model compression techniques—pruning, quantization, tensor decomposition, and knowledge distillation—for resource-constrained edge environments.