🍅🍅🍅YOLOv5-Lite: Evolved from yolov5 and the size of model is only 900+kb (int8) and 1.7M (fp16). Reach 15 FPS on the Raspberry Pi 4B~
翻译 - shufflev2-yolov5：更轻、更快、更容易部署。由yolov5进化而来，模型大小只有1.7M（int8）和3.3M（fp16）。当输入大小为320×320时，它可以在树莓派4B上达到10+ FPS~
a pytorch lib with state-of-the-art architectures, pretrained models and real-time updated results
翻译 - pytorch库，具有最新的架构，预训练的模型和实时更新的结果
#计算机科学#Light-weight Single Person Pose Estimator
This repository contains the source code of our work on designing efficient CNNs for computer vision
face image illumination quality assessment implement by pytorch
LightNet++: Boosted Light-weighted Networks for Real-time Semantic Segmentation
PyTorch implementation for 3D CNN models for medical image data (1 channel gray scale images).
Multi-backbone, Prune, Quantization, KD
QuarkDet lightweight object detection in PyTorch .Real-Time Object Detection on Mobile Devices.
#计算机科学#Tensorflow ShuffleNet v2 implementation
#计算机科学#Real-time fire detection in image/video/webcam using a convolutional neural network (deep learning) - from our ICMLA 2020 paper (Thomson / Bhowmik / Breckon)
Train/Eval the popular network by TF-Slim,include mobilenet/shufflenet/squeezenet/resnet/inception/vgg/alexnet
VoVNet, MobileNet, ShuffleNet, HarDNet, GhostNet, EfficientNet backbone networks and SKU-110K dataset for detectron2
#人脸识别#A PyTorch Implementation of ShuffleFaceNet.
implement the shufflenetv2, and test the performance
a tensorflow based implementation of ShuffleNetV2 on the Tiny ImageNet dataset
ShuffleNet_V2 for ncnn framework
Implementation of MobileNet, MobileNetv2, ShuffleNet, ShuffleNetv2, EfficientNet in Pytorch.