Thanh-Toan Do

Logic Rules Meet Deep Learning: A Novel Approach for Ship Type Classification (Extended Abstract)

Direct Quantization for Training Highly Accurate Low Bit-width Deep Neural Networks