UbiVision

NSF: CNS Core: UbiVision: Ubiquitous Machine Vision with Adaptive Wireless Networking and Edge Computing

Abstract

This project aims to realize an ambitious goal: ubiquitous machine vision (UbiVision) whose ultimate objective is to provide a platform that enables people from all over the world to share their smart cameras, which can be Uber, Airbnb, or Mobike in the context of smart cameras. The UbiVision framework consists of three main research tasks addressing challenges of dynamically managing highly coupled resources and functions across multiple technology domains: 1) camera functions such as image preprocessing and embedded machine vision; 2) network resources in the radio access network; 3) computation resources and machine vision on the edge servers. This project is interdisciplinary and integrates techniques and perspectives from wireless networking, computer vision, and edge computing in designing and optimizing UbiVision.

Principal Investigators:

• Tao Han, Ph.D. (Lead PI)
• Chen Chen, Ph.D. (PI at UCF)

Grduate Students:

• Yang Deng
• Yongjie Guan
• Ephraim Moges
• Taojiannan Yang
• Sijie Zhu
• Ce Zheng

Project Outcomes:

• Publications:

  • Xueyu Hou, Yongjie Guan, Tao Han. NeuLens: Spatial-based Dynamic Acceleration of Convolutional Neural Networks on Edge. The 28th Annual International Conference On Mobile Computing And Networking (ACM MobiCom), 2022.
  • Yongjie Guan, Xueyu Hou, Nan Wu, Bo Han, Tao Han. DeepMix: Mobility-aware, Lightweight, and Hybrid 3D Object Detection for Headsets. The 20th Annual International Conference on Mobile Systems, Applications and Services (ACM MobiSys 22), 2022.
  • Xueyu Hou, Yongjie Guan, Tao Han, Ning Zhang. DistrEdge: Speeding up Convolutional Neural Network Inference on Distributed Edge Devices. IEEE International Parallel and Distributed Processing Symposium (IPDPS), 2022.
  • Taojiannan Yang, Sijie Zhu, Matias Mendieta, Pu Wang, Ravikumar Balakrishnan, Minwoo Lee, Tao Han, Mubarak Shah, Chen Chen. MutualNet: Adaptive ConvNet via Mutual Learning from Different Model Configurations. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2021.
  • Yang Deng, Tao Han, and Nirwan Ansari, “FedVision: Federated Video Analytics with Edge Computing” in IEEE Open Journal of the Computer Society (OJCS), 2020.
  • Qiang Liu, Tao Han, Ephraim Moges, “EdgeSlice: Slicing Wireless Edge Computing Network with Decentralized Deep Reinforcement Learning ” in IEEE International Conference on Distributed Computing Systems (ICDCS), 2020. (acceptance rate: 18%)
  • Xueyu Hou and Tao Han, “TrustServing: A Quality Inspection Sampling Approach for Remote DNN Services” in IEEE International Conference on Sensing, Communication, and Networking (SECON), 2020. (accept rate: 27.9%)
  • Ephraim Moges, and Tao Han, “DecOp: Decentralized Network Operations in Software Defined Networking using Blockchain” in IEEE INFOCOM Workshop: Blockchain for Secure Software-defined Networking in Smart Communities (BlockSecSDN), 2020.
  • Sijie Zhu, Chen Chen, Waqas Sultani, “Video Anomaly Detection for Smart Surveillance”, Book Chapter of Computer Vision: A Reference Guide.
  • Taojiannan Yang, Sijie Zhu, Chen Chen, Yan Shen, Mi Zhang, Andrew Willis, “MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution”, European Conference on Computer Vision (ECCV), 2020. (Oral) (Acceptance Rate = 2% (104/5205))
  • Changlin Li, Taojiannan Yang, Sijie Zhu, Chen Chen, Shanyue Guan, “Density Map Guided Object Detection in Aerial Image”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (EarthVision Workshop), 2020. (Acceptance Rate = 26% (14/53))
  • Taojiannan Yang, Sijie Zhu, Chen Chen, “GradAug: A New Regularization Method for Deep Neural Networks”, Thirty-fourth Annual Conference on Neural Information Processing Systems (NeurIPS), 2020.

• Opensource Software:

  • EdgeSlice
  • MutualNet
  • DMNet