阅读背景：
论文翻译：BinaryNet: Training Deep Neural Networks with Weights and Activations Constrained to +1 or −1
来源：互联网
分享到：
