Efficient Table-Lookup Inference of Binarized Convolutions with Kernel-Level Binding on FPGA

Table-lookup inference, enabling convolutions via table lookups instead of multiply-accumulate (MAC) operations, shows promise for energy-efficient binarized convolutional neural networks (BCNNs). However, existing FPGA-based table-lookup BCNN designs consume considerable hardware and power due to the low binding level and LUT reuse. To address this, we propose EGSBConv, a novel convolution engine optimized for table-lookup inferences. EGSBConv leverages separable convolution and channel grouping to reduce fan-ins, facilitating kernel-level binding, while mitigating accuracy loss through internal channel extension. Each convolutional kernel is bound into a single table stored in BRAMs to reduce power consumption. Experiments on CIFAR-10 demonstrate 86 % and 94 % LUT reductions as well as <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$4.5 \times$</tex> and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$15.3 \times$</tex> power efficiency improvements compared to state-of-the-art MAC-based and table-lookup-based BCNN accelerators, respectively.