DARDA: Domain-Aware Real-Time Dynamic Neural Network Adaptation

TL;DR

DARDA introduces a proactive, resource-efficient method for real-time neural network adaptation to unseen corruptions by learning latent representations and selecting tailored sub-networks, improving performance and reducing resource use.

Contribution

The paper proposes DARDA, a novel approach that dynamically adapts neural networks to unseen corruptions using latent representations, enhancing efficiency and accuracy without extensive data.

Findings

Reduces energy consumption by 1.74x on Raspberry Pi.

Lowers cache memory footprint by 2.64x on NVIDIA Jetson Nano.

Improves classification accuracy on CIFAR-10, CIFAR-100, and TinyImagenet.

Abstract

Test Time Adaptation (TTA) has emerged as a practical solution to mitigate the performance degradation of Deep Neural Networks (DNNs) in the presence of corruption/ noise affecting inputs. Existing approaches in TTA continuously adapt the DNN, leading to excessive resource consumption and performance degradation due to accumulation of error stemming from lack of supervision. In this work, we propose Domain-Aware Real-Time Dynamic Adaptation (DARDA) to address such issues. Our key approach is to proactively learn latent representations of some corruption types, each one associated with a sub-network state tailored to correctly classify inputs affected by that corruption. After deployment, DARDA adapts the DNN to previously unseen corruptions in an unsupervised fashion by (i) estimating the latent representation of the ongoing corruption; (ii) selecting the sub-network whose associated corruption is the closest in the latent space to the ongoing corruption; and (iii) adapting DNN state, so that its representation matches the ongoing corruption. This way, DARDA is more resource efficient and can swiftly adapt to new distributions caused by different corruptions without requiring a large variety of input data. Through experiments with two popular mobile edge devices - Raspberry Pi and NVIDIA Jetson Nano - we show that DARDA reduces energy consumption and average cache memory footprint respectively by 1.74x and 2.64x with respect to the state of the art, while increasing the performance by 10.4%, 5.7% and 4.4% on CIFAR-10, CIFAR-100 and TinyImagenet.