Papers and codes about Ternary and Binary Networks for easier survey and reference. We care about the accuracy and the implementation on hardware for real-world speedup benefits.
[NeuCom 2021] Bringing AI To Edge: From Deep Learning’s Perspective
Bibtex
@article{AI_Neuro_2021,
title = {Bringing AI To Edge: From Deep Learning’s Perspective},
journal = {Neurocomputing},
year = {2021},
issn = {0925-2312},
doi = {https://doi.org/10.1016/j.neucom.2021.04.141},
url = {https://www.sciencedirect.com/science/article/pii/S0925231221016428},
author = {Di Liu and Hao Kong and Xiangzhong Luo and Weichen Liu and Ravi Subramaniam}
}[NeuCom 2021] Pruning and Quantization for Deep Neural Network Acceleration: A Survey
Bibtex
@article{PandQ_Neuro_2021,
title={Pruning and quantization for deep neural network acceleration: A survey},
author={Liang, Tailin and Glossner, John and Wang, Lei and Shi, Shaobo and Zhang, Xiaotong},
journal={Neurocomputing},
volume={461},
pages={370--403},
year={2021},
publisher={Elsevier}
}[PR 2020] [Blog] Binary Neural Networks: A Survey
Bibtex
@article{Qin:pr20_bnn_survey,
title = "Binary neural networks: A survey",
author = "Haotong Qin and Ruihao Gong and Xianglong Liu and Xiao Bai and Jingkuan Song and Nicu Sebe",
journal = "Pattern Recognition",
volume = "105",
pages = "107281",
year = "2020"
}[ArXiv 2021] A Survey of Quantization Methods for Efficient Neural Network Inference
Bibtex
@misc{gholami2021survey,
title={A Survey of Quantization Methods for Efficient Neural Network Inference},
author={Amir Gholami and Sehoon Kim and Zhen Dong and Zhewei Yao and Michael W. Mahoney and Kurt Keutzer},
year={2021},
eprint={2103.13630},
archivePrefix={arXiv},
primaryClass={cs.CV}
}[ArXiv 2018] A survey on methods and theories of quantized neural networks
Bibtex
@article{Qsuvery_ArXiv_2018,
title={A survey on methods and theories of quantized neural networks},
author={Guo, Yunhui},
journal={arXiv preprint arXiv:1808.04752},
year={2018}
}Keywords: BNN: Binary Neural Networks | TBN : Ternary-activation Binary-weight Networks| TNN: Ternary Neural Networks | mixed: Mixed Precision | INT4: 4-bit integer quantization | INT8: 8-bit integer quantization
Platforms: CPU | GPU | FPGA | ASIC | IMC: In-Memory-Computing
[CVPR_2021] [BWN+TWN] Adaptive Binary-Ternary Quantization
Bibtex
@InProceedings{BTQ_CVPR_2021,
author = {Razani, Ryan and Morin, Gregoire and Sari, Eyyub and Nia, Vahid Partovi},
title = {Adaptive Binary-Ternary Quantization},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2021},
pages = {4613-4618}
}[ECCV_2020] [BNN, QNN] ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions [PyTorch]
Bibtex
@inproceedings{liu2020reactnet,
title={ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions},
author={Liu, Zechun and Shen, Zhiqiang and Savvides, Marios and Cheng, Kwang-Ting},
booktitle={European Conference on Computer Vision (ECCV)},
year={2020}
}[CVPR_2020] [BNN, QNN] Least Squares Binary Quantization of Neural Networks [PyTorch]
Bibtex
@inproceedings{LS-BQNN_CVPR_2020,
title={Least squares binary quantization of neural networks},
author={Pouransari, Hadi and Tu, Zhucheng and Tuzel, Oncel},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
pages={698--699},
year={2020}
}[CVPR_2020] [BNN] IR-Net: Forward and Backward Information Retention for Accurate Binary Neural Networks [PyTorch]
Bibtex
@inproceedings{IR-Net_CVPR_2020,
title={Forward and backward information retention for accurate binary neural networks},
author={Qin, Haotong and Gong, Ruihao and Liu, Xianglong and Shen, Mingzhu and Wei, Ziran and Yu, Fengwei and Song, Jingkuan},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={2250--2259},
year={2020}
}[ICASSP_2020] [BNN] Balanced Binary Neural Networks with Gated Residual
Bibtex
@inproceedings{BBG_ICASSP_2020,
title={Balanced binary neural networks with gated residual},
author={Shen, Mingzhu and Liu, Xianglong and Gong, Ruihao and Han, Kai},
booktitle={ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={4197--4201},
year={2020},
organization={IEEE}
}[CVPR_2019] [BNN] CI-BCNN: Learning Channel-Wise Interactions for Binary Convolutional Neural Networks
Bibtex
@inproceedings{CI-BCNN_CVPR_2019,
title={Learning channel-wise interactions for binary convolutional neural networks},
author={Wang, Ziwei and Lu, Jiwen and Tao, Chenxin and Zhou, Jie and Tian, Qi},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
pages={568--577},
year={2019}
}[BMVC_2019] [BNN] XNOR-Net++: Improved Binary Neural Networks
Bibtex
@article{XNOR-NetPlus_BMVC_2019,
title={{XNOR-Net++}: Improved binary neural networks},
author={Bulat, Adrian and Tzimiropoulos, Georgios and Center, Samsung AI},
journal = {The British Machine Vision Conference (BMVC)},
year = {2019}
}[ECCV_2018] [BNN] Bi-Real Net: Enhancing the Performance of 1-bit CNNs with Improved Representational Capability and Advanced Training Algorithm [PyTorch, Coffee]
Bibtex
@InProceedings{Bi-Real-Net_ECCV_2018,
author = {Liu, Zechun and Wu, Baoyuan and Luo, Wenhan and Yang, Xin and Liu, Wei and Cheng, Kwang-Ting},
title = {{Bi-Real Net}: Enhancing the Performance of 1-bit CNNs with Improved Representational Capability and Advanced Training Algorithm},
booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
month = {September},
year = {2018}
}[ArXiv_2018] [BNN+CPU/GPU] BMXNet-v2: Training Competitive Binary Neural Networks from Scratch [BMXNet-v2]
Bibtex
@article{bmxnetv2,
title = {Training Competitive Binary Neural Networks from Scratch},
author = {Joseph Bethge and Marvin Bornstein and Adrian Loy and Haojin Yang and Christoph Meinel},
journal = {ArXiv e-prints},
archivePrefix = "arXiv",
eprint = {1812.01965},
Year = {2018}
}[ICM_2017] [BNN+CPU/GPU] BMXNet: An Open-Source Binary Neural Network Implementation Based on MXNet
Bibtex
@inproceedings{BMXNet_ICM_2017,
title={Bmxnet: An open-source binary neural network implementation based on mxnet},
author={Yang, Haojin and Fritzsche, Martin and Bartz, Christian and Meinel, Christoph},
booktitle={Proceedings of the 25th ACM international conference on Multimedia},
pages={1209--1212},
year={2017}
}[NeurIPS_2016] [BNN] Binarized Neural Networks
Bibtex
@incollection{BNN_NIPS_2016,
title = {Binarized Neural Networks},
author = {Hubara, Itay and Courbariaux, Matthieu and Soudry, Daniel and El-Yaniv, Ran and Bengio, Yoshua},
booktitle = {Advances in Neural Information Processing Systems 29},
pages = {4107--4115},
year = {2016},
publisher = {Curran Associates, Inc.}
}[ECCV_2016] [BNN,BWN+CPU/GPU] XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks [Torch7] [PyTorch]
Bibtex
@inproceedings{XNOR-Net_ECCV_2016,
title={Xnor-net: Imagenet classification using binary convolutional neural networks},
author={Rastegari, Mohammad and Ordonez, Vicente and Redmon, Joseph and Farhadi, Ali},
booktitle={European conference on computer vision},
pages={525--542},
year={2016},
organization={Springer}
}[AAAI_2021] [TNN] TRQ: Ternary Neural Networks With Residual Quantization
Bibtex
@inproceedings{TRQ_AAAI_2021,
title={TRQ: Ternary Neural Networks With Residual Quantization},
author={Li, Yue and Ding, Wenrui and Liu, Chunlei and Zhang, Baochang and Guo, Guodong},
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
year={2021}
}[ICCV_2021] [TNN] FATNN: Fast and Accurate Ternary Neural Networks [PyTorch]
Bibtex
@InProceedings{FATNN_ICCV_2021,
author = {Chen, Peng and Zhuang, Bohan and Shen, Chunhua},
title = {FATNN: Fast and Accurate Ternary Neural Networks},
booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
month = {October},
year = {2021},
pages = {5219-5228}
}[AAAI_2020] [TNN+FPGA] RTN: Reparameterized Ternary Network
Bibtex
@article{RTN_AAAI_2020,
title={RTN: Reparameterized Ternary Network},
author={Li, Yuhang and Dong, Xin and Zhang, Sai Qian and Bai, Haoli and Chen, Yuanpeng and Wang, Wei},
journal={The AAAI Conference on Artificial Intelligence (AAAI)},
pages={4780-4787},
year={2020}
}[NN_2018] [TNN+FPGA] GXNOR-Net: Training deep neural networks with ternary weights and activations without full-precision memory under a unified discretization framework [Theano]
Bibtex
@article{GXNOR-Net_NN_2018,
title={GXNOR-Net: Training deep neural networks with ternary weights and activations without full-precision memory under a unified discretization framework},
author={Deng, Lei and Jiao, Peng and Pei, Jing and Wu, Zhenzhi and Li, Guoqi},
journal={Neural Networks},
volume={100},
pages={49--58},
year={2018},
publisher={Elsevier}
}[ECCV_2018] [TBN] TBN: Convolutional Neural Network with Ternary Inputs and Binary Weights [PyTorch]
Bibtex
@inproceedings{TBN_ECCV_2018,
author = {Wan, Diwen and Shen, Fumin and Liu, Li and Zhu, Fan and Qin, Jie and Shao, Ling and Tao Shen, Heng},
title = {TBN: Convolutional Neural Network with Ternary Inputs and Binary Weights},
booktitle = {The European Conference on Computer Vision (ECCV)},
month = {September},
year = {2018}
}[IJCNN_2017] [TNN+FPGA] Ternary Neural Networks for Resource-Efficient AI Applications [TNN-train] [TNN-convert] [FPGA]
Bibtex
@inproceedings{TNN_IJCNN_2017,
title={Ternary neural networks for resource-efficient AI applications},
author={Alemdar, Hande and Leroy, Vincent and Prost-Boucle, Adrien and P{\'e}trot, Fr{\'e}d{\'e}ric},
booktitle={2017 international joint conference on neural networks (IJCNN)},
pages={2547--2554},
year={2017},
organization={IEEE}
}[ArXiv_2017] [TNN] FGQ-TNN: Ternary Neural Networks with Fine-Grained Quantization
Bibtex
@article{FGQ-TTN_ArXiv_2017,
title={Ternary neural networks with fine-grained quantization},
author={Mellempudi, Naveen and Kundu, Abhisek and Mudigere, Dheevatsa and Das, Dipankar and Kaul, Bharat and Dubey, Pradeep},
journal={arXiv preprint arXiv:1705.01462},
year={2017}
}[Elec_2021] [mixed] Improving Model Capacity of Quantized Networks with Conditional Computation
Bibtex
@article{CC_Elec_2021,
title={Improving Model Capacity of Quantized Networks with Conditional Computation},
author={Pham, Phuoc and Chung, Jaeyong},
journal={Electronics},
volume={10},
number={8},
pages={886},
year={2021},
publisher={Multidisciplinary Digital Publishing Institute}
}[CVPR_2021] [mixed] Layer Importance Estimation With Imprinting for Neural Network Quantization
Bibtex
@inproceedings{LIE_CVPR_2021,
title={Layer Importance Estimation With Imprinting for Neural Network Quantization},
author={Liu, Hongyang and Elkerdawy, Sara and Ray, Nilanjan and Elhoushi, Mostafa},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={2408--2417},
year={2021}
}[CVPR_2021] [mixed] AQD: Towards Accurate Quantized Object Detection [PyTorch]
Bibtex
@InProceedings{ADQ_CVPR_2021,
author = {Chen, Peng and Liu, Jing and Zhuang, Bohan and Tan, Mingkui and Shen, Chunhua},
title = {AQD: Towards Accurate Quantized Object Detection},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2021},
pages = {104-113}
}[Math_2021] [mixed] NICE: Noise Injection and Clamping Estimation for Neural Network Quantization
Bibtex
@NICE_Math_2021,
AUTHOR = {Baskin, Chaim and Zheltonozhkii, Evgenii and Rozen, Tal and Liss, Natan and Chai, Yoav and Schwartz, Eli and Giryes, Raja and Bronstein, Alexander M. and Mendelson, Avi},
TITLE = {NICE: Noise Injection and Clamping Estimation for Neural Network Quantization},
JOURNAL = {Mathematics},
VOLUME = {9},
YEAR = {2021},
NUMBER = {17},
ARTICLE-NUMBER = {2144},
URL = {https://www.mdpi.com/2227-7390/9/17/2144},
ISSN = {2227-7390},
DOI = {10.3390/math9172144}
}[ICLR_2020] [mixed] LSQ: Learned Step Size Quantization
Bibtex
@inproceedings{LSQ_ICLR_2020,
title={LEARNED STEP SIZE QUANTIZATION},
author={Steven K. Esser and Jeffrey L. McKinstry and Deepika Bablani and Rathinakumar Appuswamy and Dharmendra S. Modha},
booktitle={International Conference on Learning Representations},
year={2020},
url={https://openreview.net/forum?id=rkgO66VKDS}
}[ICLR_2020] [mixed] LLSQ: Linear Symmetric Quantization of Neural Networks for Low-precision Integer Hardware
Bibtex
@inproceedings{LLSQ_ICLR_2020,
title={Linear Symmetric Quantization of Neural Networks for Low-precision Integer Hardware},
author={Xiandong Zhao and Ying Wang and Xuyi Cai and Cheng Liu and Lei Zhang},
booktitle={International Conference on Learning Representations},
year={2020},
url={https://openreview.net/forum?id=H1lBj2VFPS}
}[CVPR_2019] [mixed] QIL: Learning to Quantize Deep Networks by Optimizing Quantization Intervals With Task Loss
Bibtex
@inproceedings{QIL_CVPR_2019,
title={Learning to quantize deep networks by optimizing quantization intervals with task loss},
author={Jung, Sangil and Son, Changyong and Lee, Seohyung and Son, Jinwoo and Han, Jae-Joon and Kwak, Youngjun and Hwang, Sung Ju and Choi, Changkyu},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={4350--4359},
year={2019}
}[ArXiv_2019] [mixed] FQ-Conv: Fully Quantized Convolution for Efficient and Accurate Inference
Bibtex
@article{FQ-Conv_Arxiv_2019,
title={FQ-conv: Fully quantized convolution for efficient and accurate inference},
author={Verhoef, Bram-Ernst and Laubeuf, Nathan and Cosemans, Stefan and Debacker, Peter and Papistas, Ioannis and Mallik, Arindam and Verkest, Diederik},
journal={arXiv preprint arXiv:1912.09356},
year={2019}
}[ECCV_2018] [mixed] LQ-Nets: Learned Quantization for Highly Accurate and Compact Deep Neural Networks [TensorPack]
Bibtex
@inproceedings{LQ-Net_ECCV_2018,
title={LQ-Nets: Learned quantization for highly accurate and compact deep neural networks},
author={Zhang, Dongqing and Yang, Jiaolong and Ye, Dongqiangzi and Hua, Gang},
booktitle={Proceedings of the European Conference on Computer Vision (ECCV)},
pages={365--382},
year={2018}
}[ArXiv_2018] [QNN] PACT: Parameterized Clipping Activation for Quantized Neural Networks [PyTorch]
Bibtex
@article{PACT_ArXiv_2018,
title={Pact: Parameterized clipping activation for quantized neural networks},
author={Choi, Jungwook and Wang, Zhuo and Venkataramani, Swagath and Chuang, Pierce I-Jen and Srinivasan, Vijayalakshmi and Gopalakrishnan, Kailash},
journal={arXiv preprint arXiv:1805.06085},
year={2018}
}[CVPR_2017] [mixed] HWGQ: Deep Learning With Low Precision by Half-Wave Gaussian Quantization
Bibtex
@InProceedings{HWGQ_CVPR_2017,
author = {Cai, Zhaowei and He, Xiaodong and Sun, Jian and Vasconcelos, Nuno},
title = {Deep Learning With Low Precision by Half-Wave Gaussian Quantization},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {July},
year = {2017}
}[ArXiv_2016] [mixed+CPU/GPU] DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients [TensorPack]
Bibtex
@article{DoReFa-Net_ArXiv_2016,
title={Dorefa-net: Training low bitwidth convolutional neural networks with low bitwidth gradients},
author={Zhou, Shuchang and Wu, Yuxin and Ni, Zekun and Zhou, Xinyu and Wen, He and Zou, Yuheng},
journal={arXiv preprint arXiv:1606.06160},
year={2016}
}[AAAI_2021] [INT8+GPU] Distribution Adaptive INT8 Quantization for Training CNNs
Bibtex
@inproceedings{DA-INT8_AAAI_2021,
title={Distribution Adaptive INT8 Quantization for Training CNNs},
author={Zhao, Kang and Huang, Sida and Pan, Pan and Li, Yinghan and Zhang, Yingya and Gu, Zhenyu and Xu, Yinghui},
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
year={2021}
}[ArXiv_2020] [INT8] Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation
Bibtex
@article{INT8_Nvidia_Arxiv_2020,
title={Integer quantization for deep learning inference: Principles and empirical evaluation},
author={Wu, Hao and Judd, Patrick and Zhang, Xiaojie and Isaev, Mikhail and Micikevicius, Paulius},
journal={arXiv preprint arXiv:2004.09602},
year={2020}
}[CVPR_2020] [INT8+GPU] UI8: Towards Unified INT8 Training for Convolutional Neural Network
Bibtex
@inproceedings{UINT8_CVPR_2020,
title={Towards unified int8 training for convolutional neural network},
author={Zhu, Feng and Gong, Ruihao and Yu, Fengwei and Liu, Xianglong and Wang, Yanfei and Li, Zhelong and Yang, Xiuqi and Yan, Junjie},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={1969--1979},
year={2020}
}[ICPADS_2020] [CPU-BNN] XOR-Net: An Efficient Computation Pipeline for Binary Neural Network Inference on Edge Devices [C++]
Bibtex
@inproceedings{XOR-Net_ICPADS_2020,
title={XOR-Net: an efficient computation pipeline for binary neural network inference on edge devices},
author={Zhu, Shien and Duong, Luan H. K. and Liu, Weichen},
booktitle={The 26th IEEE International Conference on Parallel and Distributed Systems (ICPADS)},
year={2020}
}[TVLSI_2020] [IMC-TNN] TiM-DNN: Ternary In-Memory Accelerator for Deep Neural Networks
Bibtex
@article{TiM-DNN_TVLSI_2020,
title={TiM-DNN: Ternary In-Memory Accelerator for Deep Neural Networks},
author={Jain, Shubham and Gupta, Sumeet Kumar and Raghunathan, Anand},
journal={IEEE Transactions on Very Large Scale Integration (VLSI) Systems},
year={2020},
publisher={IEEE}
}[MM_2019] [CPU-BNN, ARM] daBNN: A Super Fast Inference Framework for Binary Neural Networks on ARM devices [daBNN]
Bibtex
@inproceedings{daBNN_MM_2019,
title={dabnn: A super fast inference framework for binary neural networks on arm devices},
author={Zhang, Jianhao and Pan, Yingwei and Yao, Ting and Zhao, He and Mei, Tao},
booktitle={Proceedings of the 27th ACM International Conference on Multimedia},
pages={2272--2275},
year={2019}
}[IPDPS_2018] [CPU-BNN] BitFlow: Exploiting vector parallelism for binary neural networks on CPU
Bibtex
@inproceedings{BitFlow_2018_IPDPS,
title={Bitflow: Exploiting vector parallelism for binary neural networks on cpu},
author={Hu, Yuwei and Zhai, Jidong and Li, Dinghua and Gong, Yifan and Zhu, Yuhao and Liu, Wei and Su, Lei and Jin, Jiangming},
booktitle={2018 IEEE International Parallel and Distributed Processing Symposium (IPDPS)},
pages={244--253},
year={2018},
organization={IEEE}
}[TRETS_2018] [FPGA-TNN] High-Efficiency Convolutional Ternary Neural Networks with Custom Adder Trees and Weight Compression [FPGA]
Bibtex
@article{FPGA-TNN_TRETS_2018,
TITLE = {{High-Efficiency Convolutional Ternary Neural Networks with Custom Adder Trees and Weight Compression}},
AUTHOR = {Prost-Boucle, Adrien and BOURGE, Alban and P{\'e}trot, Fr{\'e}d{\'e}ric},
URL = {https://hal.archives-ouvertes.fr/hal-01686718},
JOURNAL = {{ACM Transactions on Reconfigurable Technology and Systems (TRETS)}},
PUBLISHER = {{ACM}},
SERIES = {Special Issue on Deep learning on FPGAs},
VOLUME = {11},
NUMBER = {3},
PAGES = {1-24},
YEAR = {2018},
MONTH = Dec,
DOI = {10.1145/3294768},
PDF = {https://hal.archives-ouvertes.fr/hal-01686718v2/file/trets_nocopyright.pdf},
HAL_ID = {hal-01686718},
HAL_VERSION = {v2},
}[FPL_2017] [FPGA-TNN] Scalable High-Performance Architecture for Convolutional Ternary Neural Networks on FPGA [FPGA]
Bibtex
@inproceedings{FPGA-TNN_FPL_2017,
title={Scalable high-performance architecture for convolutional ternary neural networks on FPGA},
author={Prost-Boucle, Adrien and Bourge, Alban and P{\'e}trot, Fr{\'e}d{\'e}ric and Alemdar, Hande and Caldwell, Nicholas and Leroy, Vincent},
booktitle={2017 27th International Conference on Field Programmable Logic and Applications (FPL)},
pages={1--7},
year={2017},
organization={IEEE}
}Bibtex