Recent searches
Search options
@TrackingActions@hachyderm.ioI’m super pumped to release #SuperAnimal models: for animal behavior, they can be zero shot for many, and if you need to fine tune, it’s 10X less data and 2X performance over the latest best pose models - please road test it!
It’s in Google Colab, so zero install needed
Blog: https://deeplabcut.medium.com/super-animal-models-are-getting-into-the-dlc-model-zoo-56962955ed2d
Colab: https://colab.research.google.com/github/DeepLabCut/DeepLabCut/blob/master/examples/COLAB/COLAB_DLC_ModelZoo.ipynb
@TrackingActions @desdemonafricker @cogneurophys #neuroethology #neuroscience This looks fantastic! Does it generate 3D pose (multi-camera inputs?)
We have loads of macaque video we’d gladly contribute. Our issue has been human bandwidth for labeling and getting 3-D pose. What’s tricky for us is that the camera captures mostly outward-facing activity (on touchscreens at the periphery).
@karihoffman @desdemonafricker @cogneurophys delighted to chat! We have some really nice 3D stuff in the pipeline, but not in SuperAnimal 