ActNAS : Generating Efficient YOLO Models using Activation NAS

TL;DR

This paper explores the use of mixed activation functions in YOLO models and introduces Activation NAS, a neural architecture search method to optimize activation function placement, resulting in faster, more memory-efficient models with improved accuracy.

Contribution

It presents a novel NAS-based approach to automatically design YOLO models with mixed activation functions for enhanced efficiency and accuracy.

Findings

Model achieved 22.28% faster inference on NPU.

Model consumed 64.15% less memory.

Slight improvement in mean Average Precision (mAP).

Abstract

Activation functions introduce non-linearity into Neural Networks, enabling them to learn complex patterns. Different activation functions vary in speed and accuracy, ranging from faster but less accurate options like ReLU to slower but more accurate functions like SiLU or SELU. Typically, same activation function is used throughout an entire model architecture. In this paper, we conduct a comprehensive study on the effects of using mixed activation functions in YOLO-based models, evaluating their impact on latency, memory usage, and accuracy across CPU, NPU, and GPU edge devices. We also propose a novel approach that leverages Neural Architecture Search (NAS) to design YOLO models with optimized mixed activation functions.The best model generated through this method demonstrates a slight improvement in mean Average Precision (mAP) compared to baseline model (SiLU), while it is 22.28% faster and consumes 64.15% less memory on the reference NPU device.