ConQ: Binary Quantization of Neural Networks via Concave Regularization | IEEE Conference Publication | IEEE Xplore