The University of Melbourne
Browse

DLC and SLEAP pre-trained exported models

Version 2 2025-06-22, 22:26
Version 1 2025-06-10, 02:30
software
posted on 2025-06-22, 22:26 authored by Leandro Aluisio ScholzLeandro Aluisio Scholz, Conrad LeeConrad Lee, Sarah StednitzSarah Stednitz, Tessa MancienneTessa Mancienne, Ethan ScottEthan Scott
<p dir="ltr">Four pre-trained zebrafish larvae pose estimation networks using DLC (ResNet-50 and ResNet-152) and SLEAP (single instance and topdown). The pose model comprises of 15 keypoints: two in each eye (anterior and posterior edges), 1 point in the swim bladder, and 10 points across the tail. The networks were trained with the same dataset of 1641 annotated frames. The exported models can be used along with Bonsai-rx or loaded within Tensorflow. <br><br>SLEAP_project_folder.zip has all files needed to load the project within SLEAP. </p>

Funding

The influence of conscious state on cortical processing and perception

Australian Research Council

Find out more...

The physics and biology of hearing in larval fish

Australian Research Council

Find out more...

Brain states and their roles in evasive behaviour

Australian Research Council

Find out more...

Assessing the structure and function of brain-wide sensory networks

National Health and Medical Research Council

Find out more...

National Institutes of Health, R01NS118406

Simons Foundation, https://ror.org/01cmst727, 625793

University of Queensland, https://ror.org/00rqy9422, RTP Scholarship

History