Cart
Create Account
Sign In
Collection:
We present a topological loss to train neural networks to segmentation fine structures with correct topology. The differentiable loss enforces the topology of the segmentation and the ground truth to be similar, based on the theory of persistent homology. The learnt network consistently outperform other methods in metrics relevant to structural accuracy. We also discuss applications of the method to other learning tasks.
- IEEE MemberUS $11.00
- Society MemberUS $0.00
- IEEE Student MemberUS $11.00
- Non-IEEE MemberUS $15.00
Videos in this product
End-To-End Training of Neural Networks with Topological Loss
We present a topological loss to train neural networks to segmentation fine structures with correct topology. The differentiable loss enforces the topology of the segmentation and the ground truth to be similar, based on the theory of persistent homology. The learnt network consistently outperform other methods in metrics relevant to structural accuracy. We also discuss applications of the method to other learning tasks.