Model HUB

On Picsellia, you willl found what we call the Model HUB which is a place where we put every available pre-trained architectures from SOTA benchmarks and where YOU can also upload the trained models you want to share with the community.

Architecture list

Here we have listed all the available architectures that we are responsible of for different frameworks.

Tensorflow 2

• efficientdet-d0

• efficientdet-d1

• efficientdet-d2

• efficientdet-d3

• efficientdet-d4

• efficientdet-d5

• faster-rcnn-inception-resnet-v2-640x640

• faster-rcnn-inception-resnet-v2-1024x1024

• faster-rcnn-resnet50-v1-640x640

• faster-rcnn-resnet50-v1-800x1333

• faster-rcnn-resnet50-v1-1024x1024

• faster-rcnn-resnet101-v1-640x640

• faster-rcnn-resnet101-v1-800x1333

• faster-rcnn-resnet101-v1-1024x1024

• faster-rcnn-resnet152-v1-640x640

• faster-rcnn-resnet152-v1-800x1333

• faster-rcnn-resnet152-v1-1024x1024

• ssd-mobilenet-v1-fpn-640

• ssd-mobilenet-v2-fpnlite-320

• ssd-mobilenet-v2-fpnlite-640

• ssd-resnet50-v1-640x640

• ssd-resnet50-v1-1024-1024

• ssd-resnet101-v1-640x640

• ssd-resnet101-v1-1024-1024

• ssd-resnet152-v1-640x640

• ssd-resnet152-v1-1024-1024

Pytorch

• YOLOv5-s

• YOLOv5-m

• YOLOv5-l

• YOLOv5-xl