Data-Free Quantization Through Weight Equalization and Bias Correction

@article{Nagel2019DataFreeQT,
  title={Data-Free Quantization Through Weight Equalization and Bias Correction},
  author={M. Nagel and Mart van Baalen and Tijmen Blankevoort and M. Welling},
  journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
  year={2019},
  pages={1325-1334}
}
  • M. Nagel, Mart van Baalen, +1 author M. Welling
  • Published 2019
  • Computer Science, Mathematics
  • 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
  • We introduce a data-free quantization method for deep neural networks that does not require fine-tuning or hyperparameter selection. [...] Key Method Our approach relies on equalizing the weight ranges in the network by making use of a scale-equivariance property of activation functions. In addition the method corrects biases in the error that are introduced during quantization. This improves quantization accuracy performance, and can be applied to many common computer vision architectures with a straight…Expand Abstract
    62 Citations
    Near-Lossless Post-Training Quantization of Deep Neural Networks via a Piecewise Linear Approximation
    • 2
    • Highly Influenced
    GQ-Net: Training Quantization-Friendly Deep Networks
    Bit Efficient Quantization for Deep Neural Networks
    • 10
    • Highly Influenced
    • PDF
    Up or Down? Adaptive Rounding for Post-Training Quantization
    • 4
    • PDF
    Improving Post Training Neural Quantization: Layer-wise Calibration and Integer Programming
    • 1
    • PDF
    Weight Equalizing Shift Scaler-Coupled Post-training Quantization
    A Statistical Framework for Low-bitwidth Training of Deep Neural Networks

    References

    SHOWING 1-10 OF 38 REFERENCES
    Fighting Quantization Bias With Bias
    • 11
    • PDF
    SYQ: Learning Symmetric Quantization for Efficient Deep Neural Networks
    • 69
    • PDF
    Quantizing deep convolutional networks for efficient inference: A whitepaper
    • 206
    • Highly Influential
    • PDF
    Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weights
    • 528
    • PDF
    Convolutional Neural Networks using Logarithmic Data Representation
    • 192
    • PDF
    Same, Same But Different - Recovering Neural Network Quantization Error Through Weight Factorization
    • 18
    • PDF
    Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference
    • 559
    • Highly Influential
    • PDF
    Variational Network Quantization
    • 39
    Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations
    • 890
    • PDF