Search papers, labs, and topics across Lattice.
1
0
2
Ditch backpropagation and per-layer clustering: FLIM-BoFP offers a faster, more efficient way to train convolutional networks using user-drawn markers and a single "bag of feature points."
