Search papers, labs, and topics across Lattice.
Affiliation:
6
0
4
Fine-grained sentiment control in text generation is now possible, with SenseShift achieving superior results over traditional decoder-based approaches.
Interactive symbolic music analysis can now seamlessly combine predictive accuracy with user-driven refinement, transforming how musicians and analysts engage with music data.
Transform audio in real-time while preserving rhythm and enhancing timbre with user-selected token grains.
Real-time music enhancement is achievable, but indiscriminate methods can actually degrade audio quality, revealing the critical need for tailored approaches.
The largest homogeneous Roman-numeral corpus ever created allows for direct note-for-note comparison across distinct analytical traditions, challenging existing standards in tonal harmony research.
Achieve improved music source restoration by cascading a BandSplit-RoFormer separator with HiFi++ GAN restorers, outperforming direct methods.
