Search papers, labs, and topics across Lattice.
1
0
3
2
Reinforcement learning can efficiently merge heterogeneous language models in federated ASR, outperforming genetic algorithms and improving character error rate.
