A Data and Compute Efficient Design for Limited-Resources Deep Learning
This work addresses the need for efficient deep learning models in medical applications on mobile devices, though it is incremental as it adapts existing methods to a specific domain.
The authors tackled the problem of making equivariant neural networks suitable for mobile devices by designing an equivariant version of MobileNetV2 and optimizing it with model quantization, achieving close-to state-of-the-art performance on the Patch Camelyon dataset while improving computational efficiency.
Thanks to their improved data efficiency, equivariant neural networks have gained increased interest in the deep learning community. They have been successfully applied in the medical domain where symmetries in the data can be effectively exploited to build more accurate and robust models. To be able to reach a much larger body of patients, mobile, on-device implementations of deep learning solutions have been developed for medical applications. However, equivariant models are commonly implemented using large and computationally expensive architectures, not suitable to run on mobile devices. In this work, we design and test an equivariant version of MobileNetV2 and further optimize it with model quantization to enable more efficient inference. We achieve close-to state of the art performance on the Patch Camelyon (PCam) medical dataset while being more computationally efficient.