This study introduces a novel progressive approximation-aware training (AAT), which efficiently integrates regularization and transfer learning techniques. The primary objective is to capture the inherent characteristics of approximate hardware designs. By considering the accuracy requirements and computational constraints inherent in the application optimizer, AAT strives to achieve an optimal balance between accuracy and power consumption. Initiating with a quantified deep neural network (DNN) model, AAT employs a range of approximation strategies to pinpoint the optimal model space and minimize energy cost. When compared to cutting-edge techniques, our approach provides remarkable energy savings, enhanced resilience against adversarial attacks, and maintains consistent accuracy.

Stochastic computing (SC) is attractive for hardware implementation due to its low complexity in arithmetic unit design; therefore, SC has attracted considerable interest to implement Artificial Neural Networks (ANNs) for resources-limited applications, because ANNs must usually perform a large number of arithmetic operations. To attain a high computation accuracy in an SC-based ANN, extended stochastic logic is utilized together with standard SC units and thus, a stochastic divider is required to perform the conversion between these logic representations. However, as the most complex SC arithmetic unit, the conventional divider incurs in a large computation latency; this limits an SC implementation for ANNs used in applications needing high performance. Therefore, there is a need to design fast stochastic dividers for SC-based ANNs. Recent works (e.g., a binary searching and triple modular redundancy (BS-TMR) based stochastic divider) are targeting a reduction in computation latency, while keeping nearly the same accuracy compared with the traditional (conventional) design. However, this divider still requires N iterations to deal with 2N-bit stochastic sequences, and thus the latency increases in proportion to the sequence length. In this paper, a decimal searching and TMR (DS-TMR) based stochastic divider is initially proposed to further reduce the computation latency; it only requires two iterations to calculate the quotient, so regardless of the sequence length. Moreover, a second trade-off design between accuracy and hardware is also presented. An SC-based Multi-Layer Perceptron (MLP) is then considered to show the effectiveness of the proposed dividers; results show that when utilizing the proposed dividers, MLP achieves the lowest computation latency while keeping the classification results at the same accuracy. When using as combined metric the product of the latency and power dissipation, the proposed designs are also shown to be superior to the SC-based MLPs employing other dividers found in the technical literature as well as the commonly used 32-bit floating point implementation. This makes the proposed dividers very attractive compared with the existing schemes for SC-based ANNs.