BasisNet: Two-stage Model Synthesis for Efficient Inference

TL;DR

BasisNet introduces a two-stage neural network synthesis approach combining lightweight preview models and specialist models, achieving high accuracy with reduced computational cost on ImageNet.

Contribution

It proposes a novel two-stage model synthesis method that enhances efficiency and accuracy in neural networks through joint training and input-dependent model combination.

Findings

Achieved 80.3% top-1 accuracy with 290M MAdds on ImageNet.

Halved the computational cost compared to previous state-of-the-art.

Further reduced average cost to 198M MAdds with early termination while maintaining 80.0% accuracy.

Abstract

In this work, we present BasisNet which combines recent advancements in efficient neural network architectures, conditional computation, and early termination in a simple new form. Our approach incorporates a lightweight model to preview the input and generate input-dependent combination coefficients, which later controls the synthesis of a more accurate specialist model to make final prediction. The two-stage model synthesis strategy can be applied to any network architectures and both stages are jointly trained. We also show that proper training recipes are critical for increasing generalizability for such high capacity neural networks. On ImageNet classification benchmark, our BasisNet with MobileNets as backbone demonstrated clear advantage on accuracy-efficiency trade-off over several strong baselines. Specifically, BasisNet-MobileNetV3 obtained 80.3% top-1 accuracy with only 290M Multiply-Add operations, halving the computational cost of previous state-of-the-art without sacrificing accuracy. With early termination, the average cost can be further reduced to 198M MAdds while maintaining accuracy of 80.0% on ImageNet.