Publications by authors named "Vijay Venu Thiyagarajan"
Article Synopsis
- - Producing accurate 3D models from biological images, especially of complex brain structures, requires extensive human effort to annotate data, which is time-consuming and typically done by experts.
- - The authors developed a new deep learning method that allows for quick 3D segmentations using minimal 2D annotations, dramatically reducing the time required to create training data.
- - This innovative approach enables non-experts to generate necessary annotations efficiently, making it easier to study brain circuits and their connections across larger datasets.
Article Synopsis
- Producing dense 3D reconstructions from biological imaging is tough and usually needs a lot of precise training data for deep learning models.
- The brain's neuropil, with its complex structure of various cellular processes, is particularly challenging to annotate with traditional methods.
- We created a new deep learning approach that allows for quick 3D segmentations from minimal 2D annotations, reducing human annotation time significantly and enabling non-experts to contribute effectively.
