<monospace>CamoNet</monospace>: On-Device Neural Network Adaptation With Zero Interaction and Unlabeled Data for Diverse Edge Environments

Deploying deep learning models to edge devices for low-latency and privacy-preserving applications has become a trend. To adapt to heterogeneous devices and data, it is significant to generate customized models. However, existing model adaptation approaches require edge devices to make interactions (collecting hardware information or local data) with the cloud, which raises privacy concerns, increases communication costs, and burdens the cloud. By contrast, we propose CamoNet, a universal on-device model adaptation framework with zero interaction between devices and the cloud. In CamoNet, a lightweight on-device neural architecture search module is utilized to quickly generate a customized model for subsequent on-device training, followed by an on-device contrastive transfer learning module to effectively leverage unlabeled data for fine-tuning the customized model. Extensive experimental results show that CamoNet can effectively run on various edge devices. Compared with the SOTA model adaptation approaches, CamoNet achieves significant accuracy improvement by 25.2% on average for image classification, 10.1% on average for object detection, and reduces the training memory by 4.8-11.4&#215;. We will open-source our models and tools for edge AI developers.