RepAct: The Re-parameterizable Adaptive Activation Function
This work addresses the need for efficient AI in IoT and edge computing, offering an incremental improvement in activation functions for lightweight networks.
The paper tackled the problem of optimizing lightweight neural networks for IoT and edge computing by introducing RepAct, a re-parameterizable adaptive activation function, which achieved up to a 7.92% accuracy boost on MobileNetV3-Small for ImageNet100 while maintaining computational efficiency.
Addressing the imperative need for efficient artificial intelligence in IoT and edge computing, this study presents RepAct, a re-parameterizable adaptive activation function tailored for optimizing lightweight neural networks within the computational limitations of edge devices. By employing a multi-branch structure with learnable adaptive weights, RepAct enriches feature processing and enhances cross-layer interpretability. When evaluated on tasks such as image classification and object detection, RepAct notably surpassed conventional activation functions in lightweight networks, delivering up to a 7.92% accuracy boost on MobileNetV3-Small for the ImageNet100 dataset, while maintaining computational complexity on par with HardSwish. This innovative approach not only maximizes model parameter efficiency but also significantly improves the performance and understanding capabilities of lightweight neural networks, demonstrating its potential for real-time edge computing applications.