Search papers, labs, and topics across Lattice.
3
0
4
Training-free diffusion models can now harmonize satellite imagery across diverse domains, enabling scalable remote-sensing synthesis without retraining.
Unlock progress in child safety AI: a new dataset lets researchers train models on the structure of abusive images, without ever seeing the original content.
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."
