Kai Li (李凯)

A curated leaderboard that ranks the audio field three ways at once, and reveals why the strongest projects now win on research, demo quality, and open-source execution together.

A causal separator from Tsinghua that swaps full hindsight for a rolling memory, and tries to erase the usual quality penalty of live audio.

A synthetic audio stack that treats motion as the signal, not metadata, then turns that motion into a dataset and benchmark for speech enhancement and separation.

A Markdown skill that turns arXiv monitoring into a checkpointed review pipeline, with a skeptical reviewer persona, recovery-safe caching, and doc publishing baked in.

A benchmark that scores audio systems on the failures text-only evals miss: spoofing, noise, accent bias, privacy leaks, hallucination, and safety drift.

Apollo-data-preprocess is not a model repo. It is the filter that strips silence, slices tracks into learnable windows, and turns raw stems into HDF5 training fuel for high-fidelity music restoration.

A real-time audio-visual model that uses mouth motion, causal convolutions, and SRUs to pull one voice out of a crowd without peeking into the future.

This repo is the training scaffolding behind Gull, a generative audio codec that compresses and reconstructs sound from subbands instead of raw waveforms.

A look at how this Tsinghua-built repo uses time-frequency interleaving and gain extraction to chase strong separation without the usual model bloat.

A terminal-style site, a Git-backed CMS, and an MCP server combine so an AI agent can help build and maintain a résumé-to-portfolio engine.

A one-file userscript turns Scholar’s citation maze into a background lookup, then hunts for the most canonical BibTeX entry before you ever open the popup.

How structure-aware "pages" and dual-path pooling are replacing token-level pruning to solve the LLM lost-in-the-middle problem.