온디바이스 비전 검사를 위한 경량 AI 프레임워크 성능 분석 = Performance Analysis of Lightweight AI Frameworks for On-Device Vision Inspection

다국어 초록 (Multilingual Abstract)

This paper analyzes the performance of lightweight AI frameworks in edge computing environments, addressing the growing need for on-device vision inspection due to its benefits in real-time processing, data privacy, and reduced network latency. The study compares the performance of almond defect detection models based on MobileNetV2 and MobileNetV3 Small, implemented using lightweight PyTorch and TensorFlow Lite on a Raspberry Pi 4 Model B. The main evaluation metrics include model file size, memory usage, inference time, accuracy, precision, recall, and F1-score. Results indicate that TensorFlow Lite with post-training quantization generally achieves the smallest model size and fastest inference speed. However, lightweight PyTorch models demonstrate superior accuracy. The study demonstrates the potential for industrial applications of edge AI in vision inspection tasks and emphasizes the need for further research on various hardware platforms, optimization techniques, and lightweight model architectures to enhance the efficiency and effectiveness of edge AI solutions.

참고문헌 (Reference)

1 "https://www.kaggle.com/datasets/mahyeks/almond-prunus-du lcis-damage-detection"

2 "https://www.image-net.org/"

3 M. Satyanarayanan, "The Emergence of Edge Computing" 50 (50): 30-39, 2017

4 L. Ye, "The Challenges and Emerging Technologies for Low-Power Artificial Intelligence IoT Systems" 68 (68): 4821-4834, 2021

5 A. Howard, "Searching for MobileNetv3" 1314-1324, 2019

6 C. Profentzas, "Performance of Deep Neural Networks on Low-power IoT Devices" 32-37, 2021

7 W. Glegoła, "MobileNet family tailored for Raspberry Pi" 192 : 2249-2258, 2021

8 "Image Classification with MobilenetV2, Arm NN, and TensorFlow Lite Delegate pre-built binaries, Version 21.11Tutorial"

9 M. Yurdakul, "Flower Pollination Algorithm-Optimized Deep CNN Features for Almond(Prunus dulcis) Classification" 433-438, 2024

10 J. Ryu, "Experimental Analysis of The MobileNetV2 Lightweight Method for Leveraging Deep Learning Model in Mobile Devices" 633-635, 2020

11 Z. Zhou, "Edge Intelligence : Paving the Last Mile of Artificial Intelligence with Edge Computing" 107 (107): 1738-1762, 2019

12 W. Shi, "Edge Computing : Vision and Challenges" 3 (3): 637-646, 2016

13 E. Li, "Edge AI : On-Demand Accelerating Deep Neural Network Inference via Edge Computing" 19 (19): 447-457, 2020

14 F. MohiEldeen Alabbasya, "Compressing Medical Deep Neural Network Models for Edge Devices Using Knowledge Distillation" 35 (35): 101616-, 2023

15 C. Luo, "Comparison and Benchmarking of AI Models and Frameworks on Mobile Devices"

16 Y. Zhao, "Automatic Recognition of Surface Defects of Hot Rolled Strip Steel Based on Deep Parallel Attention Convolution Neural Network" 353 : 135313-, 2023

17 Aparna, "An Overview of Industrial Vision Systems" 3 (3): 201-207, 2017