tokenization

1 article

sort: new top best

bug-bounty524 xss351 exploit241 google229 rce184 facebook171 microsoft146 web3122 writeup114 malware111 cve107 open-source91 csrf86 apple82 account-takeover80 sqli68 browser64 ai-agents63 cloudflare60 dos60 phishing58 privilege-escalation53 ssrf52 supply-chain47 tool46 pentest46 reverse-engineering46 privacy44 auth-bypass43 idor38 oauth38 llm37 aws37 ctf36 cloud36 opinion35 race-condition34 automation33 lfi32 machine-learning32 infrastructure31 code-generation31 node30 clickjacking29 cors28 access-control27 react26 subdomain-takeover26 info-disclosure25 rust24

0 2/10

TADA: Speech generation through text-acoustic synchronization

research

Hume AI open-sources TADA, a novel text-to-speech system using one-to-one text-acoustic token alignment that achieves 5x faster inference than competing LLM-based TTS systems while eliminating hallucinations and enabling on-device deployment. The approach synchronizes text and speech through a dual alignment tokenization schema, reducing the token mismatch problem inherent in traditional audio representation methods.

text-to-speech speech-synthesis llm tokenization audio-processing machine-learning on-device-deployment open-source

TADA Hume AI Llama LibriTTSR EARS dataset HuggingFace

hume.ai · smusamashah · 5 days ago · details · hn