pith:6PATNCQZ
Unifying Speech Editing Detection and Content Localization via Prior-Enhanced Audio LLMs
Audio LLMs reformulated as text generators detect speech edits and localize their content on a new realistic dataset.
arxiv:2601.21463 v3 · 2026-01-29 · cs.SD · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{6PATNCQZCHQ6OJI33H2LKXGZOS}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
Experimental results demonstrate that the proposed approach consistently outperforms existing methods across both detection and localization tasks.
That the AiEdit dataset constructed with state-of-the-art end-to-end speech editing systems provides realistic coverage of modern editing threats and that the prior-enhanced prompting strategy successfully grounds the generative model in acoustic evidence.
The paper introduces the AiEdit dataset of 140 hours of realistic speech edits and a prior-enhanced Audio LLM framework that unifies detection and localization via text generation and an acoustic consistency loss.
Cited by
Receipt and verification
| First computed | 2026-05-26T01:03:23.487775Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
f3c1368a1911e1e7251bd9f4b55cd974b6bf184126a2b23260df221e6b53f4a4
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/6PATNCQZCHQ6OJI33H2LKXGZOS \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: f3c1368a1911e1e7251bd9f4b55cd974b6bf184126a2b23260df221e6b53f4a4
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "5af8683e7ca58e77d82192569bac84d26da273c8b0fcdf206a3a5dda6ff6eebd",
"cross_cats_sorted": [
"cs.AI"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"primary_cat": "cs.SD",
"submitted_at": "2026-01-29T09:39:28Z",
"title_canon_sha256": "b894be1e20f9458f07b5150de3e1fdf599ffe125ef2f27f64edad3fe52c789b2"
},
"schema_version": "1.0",
"source": {
"id": "2601.21463",
"kind": "arxiv",
"version": 3
}
}