Calibration-Aided Edge Inference Offloading via Adaptive Model Partitioning of Deep Neural Networks
This work addresses the challenge of efficient edge-cloud inference offloading for mobile devices, but it is incremental as it builds on existing early-exit and calibration methods.
The paper tackles the problem of communication delay in offloading DNN inference from mobile devices to the cloud by using adaptive model partitioning with early-exit DNNs, and shows that miscalibration in these DNNs can significantly decrease inference accuracy, which can be solved by implementing a calibration algorithm prior to deployment.
Mobile devices can offload deep neural network (DNN)-based inference to the cloud, overcoming local hardware and energy limitations. However, offloading adds communication delay, thus increasing the overall inference time, and hence it should be used only when needed. An approach to address this problem consists of the use of adaptive model partitioning based on early-exit DNNs. Accordingly, the inference starts at the mobile device, and an intermediate layer estimates the accuracy: If the estimated accuracy is sufficient, the device takes the inference decision; Otherwise, the remaining layers of the DNN run at the cloud. Thus, the device offloads the inference to the cloud only if it cannot classify a sample with high confidence. This offloading requires a correct accuracy prediction at the device. Nevertheless, DNNs are typically miscalibrated, providing overconfident decisions. This work shows that the employment of a miscalibrated early-exit DNN for offloading via model partitioning can significantly decrease inference accuracy. In contrast, we argue that implementing a calibration algorithm prior to deployment can solve this problem, allowing for more reliable offloading decisions.