Dynamic Slicing for Deep Neural Networks

被引:19
|
作者
Zhang, Ziqi [1 ]
Li, Yuanchun [2 ]
Guo, Yao [1 ]
Chen, Xiangqun [1 ]
Liu, Yunxin [2 ]
机构
[1] Peking Univ, Dept Comp Sci, Key Lab High Confidence Software Technol MOE, Beijing, Peoples R China
[2] Microsoft Res, Beijing, Peoples R China
基金
中国国家自然科学基金;
关键词
Program slicing; deep neural networks; dynamic slicing; data flow analysis;
D O I
10.1145/3368089.3409676
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Program slicing has been widely applied in a variety of software engineering tasks. However, existing program slicing techniques only deal with traditional programs that are constructed with instructions and variables, rather than neural networks that are composed of neurons and synapses. In this paper, we propose NNSlicer, the first approach for slicing deep neural networks based on data flow analysis. Our method understands the reaction of each neuron to an input based on the difference between its behavior activated by the input and the average behavior over the whole dataset. Then we quantify the neuron contributions to the slicing criterion by recursively backtracking from the output neurons, and calculate the slice as the neurons and the synapses with larger contributions. We demonstrate the usefulness and effectiveness of NNSlicer with three applications, including adversarial input detection, model pruning, and selective model protection. In all applications, NNSlicer significantly outperforms other baselines that do not rely on data flow analysis.
引用
收藏
页码:838 / 850
页数:13
相关论文
共 50 条
  • [1] Deep Reinforcement Learning for Dynamic Network Slicing in IEEE 802.11 Networks
    De Bast, Sibren
    Torrea-Duran, Rodolfo
    Chiumento, Alessandro
    Pollin, Sofie
    Gacanin, Haris
    [J]. IEEE CONFERENCE ON COMPUTER COMMUNICATIONS WORKSHOPS (IEEE INFOCOM 2019 WKSHPS), 2019, : 264 - 269
  • [2] Human Action Classification Using Temporal Slicing for Deep Convolutional Neural Networks
    Henderson, Nathan
    Aygun, Ramazan
    [J]. 2017 IEEE INTERNATIONAL SYMPOSIUM ON MULTIMEDIA (ISM), 2017, : 83 - 90
  • [3] DeepDyve: Dynamic Verification for Deep Neural Networks
    Li, Yu
    Li, Min
    Luo, Bo
    Tian, Ye
    Xu, Qiang
    [J]. CCS '20: PROCEEDINGS OF THE 2020 ACM SIGSAC CONFERENCE ON COMPUTER AND COMMUNICATIONS SECURITY, 2020, : 101 - 112
  • [4] Fully Dynamic Inference With Deep Neural Networks
    Xia, Wenhan
    Yin, Hongxu
    Dai, Xiaoliang
    Jha, Niraj K.
    [J]. IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING, 2022, 10 (02) : 962 - 972
  • [5] Optimal and Fast Real-Time Resource Slicing With Deep Dueling Neural Networks
    Nguyen Van Huynh
    Dinh Thai Hoang
    Nguyen, Diep N.
    Dutkiewicz, Eryk
    [J]. IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, 2019, 37 (06) : 1455 - 1470
  • [6] Crowding Game and Deep Q-Networks for Dynamic RAN Slicing in 5G Networks
    Saad, Joe
    Khawam, Kinda
    Yassin, Mohamad
    Costanzo, Salvatore
    Boulos, Karen
    [J]. PROCEEDINGS OF THE 20TH ACM INTERNATIONAL SYMPOSIUM ON MOBILITY MANAGEMENT AND WIRELESS ACCESS, MOBIWAC 2022, 2022, : 37 - 46
  • [7] Dynamic Data Fault Localization for Deep Neural Networks
    Yin, Yining
    Feng, Yang
    Weng, Shihao
    Liu, Zixi
    Yao, Yuan
    Zhang, Yichi
    Zhao, Zhihong
    Chen, Zhenyu
    [J]. PROCEEDINGS OF THE 31ST ACM JOINT MEETING EUROPEAN SOFTWARE ENGINEERING CONFERENCE AND SYMPOSIUM ON THE FOUNDATIONS OF SOFTWARE ENGINEERING, ESEC/FSE 2023, 2023, : 1345 - 1357
  • [8] Deep Dynamic Neural Networks for Gesture Segmentation and Recognition
    Wu, Di
    Shao, Ling
    [J]. COMPUTER VISION - ECCV 2014 WORKSHOPS, PT I, 2015, 8925 : 552 - 571
  • [9] Dynamic Graph Segmentation for Deep Graph Neural Networks
    Kang, Johan Kok Zhi
    Yang, Suwei
    Venkatesan, Suriya
    Tan, Sien Yi
    Cheng, Feng
    He, Bingsheng
    [J]. PROCEEDINGS OF THE 28TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, KDD 2022, 2022, : 4601 - 4611
  • [10] DYVEDEEP: Dynamic Variable Effort Deep Neural Networks
    Ganapathy, Sanjay
    Venkataramani, Swagath
    Sriraman, Giridhur
    Ravindran, Balaraman
    Raghunathan, Anand
    [J]. ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS, 2020, 19 (03)