Optics
量化感知训练
Quantization-aware trainingA training technique that simulates the effects of quantization during the forward and backward passes of a neural network to improve model accuracy and robustness under hardware constraints. / 一种在神经网络的前向和反向传播过程中模拟量化效应的训练技术，以提高模型在硬件约束下的准确性和鲁棒性。
- Knowledge type
- 工艺技术
- First proposed or discovered
- Proposer
- Turkey Farm member
- Creator
- 农场主
- Last updated
- 2026-11-04 17:05:41
Sub-knowledge points
Circular total-variation (TV) regularizationKerr nonlinear phase modulationPerturbation-aware training
Higher-order knowledge points
Diffractive neural networksAngular-spectrum propagation model
Main parameters and performance indicators
bit-width: The number of bits used to represent weights and activations (e.g., 8-level phase deployment).non-linear phase modulation: Kerr effect used to introduce non-linear degrees of freedom.
Open problems
No unresolved problem recorded yet
Related literature
ranked by community starsQuantization-aware nonlinear diffractive neural network for robust optical image reconstruction under fabrication constraintsOpen source
Related products and links
No item recorded yet
Creation and modification trail
1 recordsRecord title: Quantization-aware training.
Node update feedback
review required