Lets Keep it simple, Using simple architectures to outperform deeper and more complex architectures

Wide Residual Networks

Award winning ConvNets from 2014 ImageNet ILSVRC challenge

Alexnet-level accuracy with 50x fewer parameters.

Brain-inspired Multilayer Perceptron with Spiking Neurons

An efficient ConvNet optimized for speed and memory, pre-trained on ImageNet

Next generation ResNets, more efficient and accurate

Deep residual networks pre-trained on ImageNet

Efficient networks optimized for speed and memory, with residual blocks

Also called GoogleNetv3, a famous ConvNet trained on ImageNet from 2015

Efficient networks by generating more features from cheap operations

Fully-Convolutional Network model with ResNet-50 and ResNet-101 backbones

Dense Convolutional Network (DenseNet), connects each layer to every other layer in a feed-forward fashion.

The 2012 ImageNet winner achieved a top-5 error of 15.3%, more than 10.8 percentage points…

DeepLabV3 models with ResNet-50, ResNet-101 and MobileNet-V3 backbones

Ultralytics YOLOv5 🚀 for object detection, instance segmentation and image classification.