efficient-inference

Efficient AI Backbones including GhostNet, TNT and MLP, developed by Huawei Noah's Ark Lab.

Code for paper " AdderNet: Do We Really Need Multiplications in Deep Learning?"

Learning Efficient Convolutional Networks through Network Slimming, In ICCV 2017.

[CVPR 2021] Exploring Sparsity in Image Super-Resolution for Efficient Inference

Deep Face Model Compression

(CVPR 2021, Oral) Dynamic Slimmable Network

Soft Threshold Weight Reparameterization for Learnable Sparsity

This is an official implementation for "DeciWatch: A Simple Baseline for 10x Efficient 2D and 3D Pose Estimation" (ECCV 2022)

Implementation of AAAI 21 paper: Nested Named Entity Recognition with Partially Observed TreeCRFs

Jia-Hong Lee, Yi-Ming Chan, Ting-Yen Chen, and Chu-Song Chen, "Joint Estimation of Age and Gender from Unconstrained Face Images using Lightweight Multi-task CNN for Mobile Applications," IEEE International Conference on Multimedia Information Processing and Retrieval, MIPR 2018

[ECCV 2020] Code release for "Resolution Switchable Networks for Runtime Efficient Image Recognition"

Official implementation of AdaMML. https://arxiv.org/abs/2105.05165.

Concise, Modular, Human-friendly PyTorch implementation of EfficientNet with Pre-trained Weights.

[ICLR 2020] ”Triple Wins: Boosting Accuracy, Robustness and Efficiency Together by Enabling Input-Adaptive Inference“

Yi-Min Chou, Yi-Ming Chan, Jia-Hong Lee, Chih-Yi Chiu, Chu-Song Chen, "Unifying and Merging Well-trained Deep Neural Networks for Inference Stage," International Joint Conference on Artificial Intelligence (IJCAI), 2018

Code for paper: 'Multi-Component Optimization and Efficient Deployment of Neural-Networks on Resource-Constrained IoT Hardware'

Stochastic Downsampling for Cost-Adjustable Inference and Improved Regularization in Convolutional Networks

[FPGA'21] CoDeNet is an efficient object detection model on PyTorch, with SOTA performance on VOC and COCO based on CenterNet and Co-Designed deformable convolution.

Supplementary material for IEEE Services Computing paper title 'An SRAM Optimized Approach for Constant Memory Consumption and Ultra-fast Execution of ML Classifiers on TinyML Hardware'

Cheng-Hao Tu, Jia-Hong Lee, Yi-Ming Chan and Chu-Song Chen, "Pruning Depthwise Separable Convolutions for MobileNet Compression," International Joint Conference on Neural Networks, IJCNN 2020, July 2020.

Cheng-En Wu, Yi-Ming Chan and Chu-Song Chen "On Merging MobileNets for Efficient Multitask Inference", International Symposium on High-Performance Computer Architecture(HPCA) on Workshop on Energy Efficient Machine Learning and Cognitive Computing for Embedded Applications(EMC2), 2019

Repo of paper: TinyML Benchmark: Executing Fully Connected Neural Networks on Commodity Microcontrollers

List of papers related to neural network quantization in recent AI conferences and journals (please refer to README in the folder for a better view).

Finding Storage- and Compute-Efficient Convolutional Neural Networks

Extremely light-weight MixNet with Top-1 75.7% and 2.5M params

NeurIPS 2019 MicroNet Challenge

BLOCKSET: Efficient out of core tree ensemble inference

[MicroNet Challenge (NeurIPS 2019 )] "Adjustable Quantization: Jointly Learn the Bit-width and Weight in DNN Training" by Yonggan Fu, Ruiyang Zhao, Yue Wang, Chaojian Li, Haoran You, Zhangyang Wang, Yingyan Lin

Code for paper: 'Edge2Train: a framework to train machine learning models (SVMs) on resource-constrained IoT edge devices'

Exploring Variational Deep Q Networks. A study undertaken for the University of Cambridge's R244 Computer Science Masters Course. Inspired by https://arxiv.org/abs/1711.11225/.