Energy-Constrained Compression for Deep Neural Networks via Weighted Sparse Projection and Layer Input Masking
This addresses energy efficiency for DNNs in resource-limited devices, representing an incremental improvement over existing energy-saving techniques.
The paper tackles the problem of reducing energy consumption in deep neural networks for deployment in constrained environments like autonomous drones, proposing an end-to-end training framework with quantitative energy guarantees via weighted sparse projection and input masking, resulting in higher accuracies under same or lower energy budgets compared to prior methods.
Deep Neural Networks (DNNs) are increasingly deployed in highly energy-constrained environments such as autonomous drones and wearable devices while at the same time must operate in real-time. Therefore, reducing the energy consumption has become a major design consideration in DNN training. This paper proposes the first end-to-end DNN training framework that provides quantitative energy consumption guarantees via weighted sparse projection and input masking. The key idea is to formulate the DNN training as an optimization problem in which the energy budget imposes a previously unconsidered optimization constraint. We integrate the quantitative DNN energy estimation into the DNN training process to assist the constrained optimization. We prove that an approximate algorithm can be used to efficiently solve the optimization problem. Compared to the best prior energy-saving methods, our framework trains DNNs that provide higher accuracies under same or lower energy budgets. Code is publicly available.