lfranke/VET

Custom dataset

Closed this issue · 0 comments

Hi,

I get an issue when training with a custom dataset.

cropped sampler 1400x2104 to 512 x 512 render scale 1 cropped sampler 1400x2104 to 512 x 512 render scale 1 full sampler 1400x2104 render scale 1 full sampler 1400x2104 render scale 1 POINT GRADIENTS ARE COMPUTED. Assertion 'n.allFinite()' failed! File: ./VET/src/lib/rendering/NeuralPointCloudCuda.cpp:33 Function: NeuralPointCloudCudaImpl::NeuralPointCloudCudaImpl(const Saiga::UnifiedMesh&, bool, float, Saiga::AABB) Abortado (core' generado)
`

I've cheked the .ply for NaN values and checked for normalization.

Thank you.