Zhangzhe's Blog

08-22

Effective Quantization Methods for Recurrent Neural Networks

08-22

Deep Learning with Low Precision by Half-wave Gaussian Quantization

08-22

Two-Step Quantization for Low-bit Neural Networks

08-22

DO-Conv: Depthwise Over-parameterized Convolutional Layer

08-22

PSConv: Squeezing Feature Pyramid into One Compact Poly-Scale Convolutional Layer

08-17

Diverse Branch Block: Building a Convolution as an Inception-like Unit

08-17

MLP-Mixer: An all-MLP Architecture for Vision

08-17

RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition

08-17

CondConv: Conditionally Parameterized Convolutions for Efficient Inference

08-01

ULSAM: Ultra-Lightweight Subspace Attention Module for Compact Convolutional Neural Networks