MCPcopy Create free account
hub / github.com/chrischoy/3D-R2N2 / BlockDiagonalLayer

Class BlockDiagonalLayer

lib/layers.py:164–203  ·  view source on GitHub ↗

Compute block diagonal matrix multiplication efficiently using broadcasting Last dimension will be used for matrix multiplication. prev_layer.output_shape = N x D_1 x D_2 x ... x D_{n-1} x D_n output_shape = N x D_1 x D_2 x ... x D_{n-1} x n_out

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers

nothing calls this directly

Calls

no outgoing calls

Tested by

no test coverage detected