feb 9, 2016 - M Courbariaux:
Binarized neural networks:
Training deep neural networks
with weights and activations
constrained to+ 1 or-1 (BNNs)
Description:
BNN被提出,只使用+1和-1实现权重和激活值的量化,但是依赖于BN层,并且精度下降较多。
Added to timeline:
Quantization timeline
The quantization development history
Date: