R-TOSS: A Framework for Real-Time Object Detection using Semi-Structured Pruning

被引：5

作者：

Balasubramaniam, Abhishek ^{[1
]}

Sunny, Febin ^{[1
]}

Pasricha, Sudeep ^{[1
]}

机构：

[1] Colorado State Univ, Dept Elect & Comp Engn, Ft Collins, CO 80523 USA

来源：

2023 60TH ACM/IEEE DESIGN AUTOMATION CONFERENCE, DAC | 2023年

关键词：

pruning; object detection; YOLOv5; RetinaNet; Jetson TX2; model compression; computer vision;

D O I：

10.1109/DAC56929.2023.10247917

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Object detectors used in autonomous vehicles can have high memory and computational overheads. In this paper, we introduce a novel semi-structured pruning framework called R-TOSS that overcomes the shortcomings of state-of-the-art model pruning techniques. Experimental results on the JetsonTX2 platform show that R-TOSS has a compression rate of 4.4x on the YOLOv5 object detector with a 2.15x speedup in inference time and 57.01% decrease in energy usage. R-TOSS also enables 2.89x compression on RetinaNet with a 1.86x speedup in inference time and 56.31% decrease in energy usage. We also demonstrate significant improvements compared to various state-of-the-art pruning techniques.

引用

页数：6

共 50 条

[21] SODA: A Real-time Simulation Framework for Object Detection and Analysis in Smart Manufacturing [J].

Lasek, Piotr .

KNOWLEDGE-BASED AND INTELLIGENT INFORMATION & ENGINEERING SYSTEMS (KSE 2021), 2021, 192 :923-930

[22] Object detection based on color and shape features for service robot in semi-structured indoor environment [J].

Li, Haojie ;

Zhao, Qijie ;

Li, Xianfa ;

Zhang, Xudong .

INTERNATIONAL JOURNAL OF INTELLIGENT ROBOTICS AND APPLICATIONS, 2019, 3 (04) :430-442

[23] Object detection based on color and shape features for service robot in semi-structured indoor environment [J].

Haojie Li ;

Qijie Zhao ;

Xianfa Li ;

Xudong Zhang .

International Journal of Intelligent Robotics and Applications, 2019, 3 :430-442

[24] Real-Time Indoor Path Planning Using Object Detection for Autonomous Flying Robots [J].

Alparslan, Onder ;

Cetin, Omer .

INTELLIGENT AUTOMATION AND SOFT COMPUTING, 2023, 36 (03) :3355-3370

[25] Region Boosting for Real-Time Object Detection Using Multi-Dimensional Attention [J].

Chen, Jinlong ;

Xu, Kejian ;

Ning, Yi ;

Xu, Zhi .

IEEE ACCESS, 2024, 12 :171634-171643

[26] Real-time Object Detection Based on R-FCN Network Under Structured Scene of High-speed Railway [J].

Han, Qian ;

Wang, Shengchun ;

Gu, Zichen ;

Dai, Peng ;

Feng, Qibo .

TWELFTH INTERNATIONAL CONFERENCE ON DIGITAL IMAGE PROCESSING (ICDIP 2020), 2020, 11519

[27] Real-Time Adaptive Object Detection and Tracking for Autonomous Vehicles [J].

Hoffmann, Joao Eduardo ;

Tosso, Hilkija Gaius ;

Dias Santos, Max Mauro ;

Justo, Joao Francisco ;

Malik, Asad Waqar ;

Rahman, Anis Ur .

IEEE TRANSACTIONS ON INTELLIGENT VEHICLES, 2021, 6 (03) :450-459

[28] An Embedded Real-Time Object Detection and Measurement of its Size [J].

Othman, Nashwan Adnan ;

Salur, Mehmet Umut ;

Karakose, Mehmet ;

Aydin, Ilhan .

2018 INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND DATA PROCESSING (IDAP), 2018,

[29] Real-time object detection applied on drones [J].

Wei, Jingjing ;

Zhao, Yiding .

International Agricultural Engineering Journal, 2019, 28 (04) :450-459

[30] Real-time Single Object Detection on The UAV [J].

Wu, Hsiang-Huang ;

Zhou, Zejian ;

Feng, Ming ;

Yan, Yuzhong ;

Xu, Hao ;

Qian, Lijun .

2019 INTERNATIONAL CONFERENCE ON UNMANNED AIRCRAFT SYSTEMS (ICUAS' 19), 2019, :1013-1022

← 1 2 3 4 5 →