One hundred researchers, 394 on-topic titles: Wang, X. leads audio-deepfake detection at 18, and only 39 of 100 surface in public-facing media

Asked:

“Who are the top 100 researchers by both scientific and non scientific volume on the topic of deepfake audio detection and scam audio detection”

A live-web bibliometric scan across 12 query families retrieved 1,849 unique source pages and 8,370 author–work records; after title-level topic filtering, 100 researchers were ranked lexicographically by scientific volume (distinct on-topic scholarly titles per researcher), then verified non-scientific volume, then corroborating scholarly pages. Counts are researcher-level, so a coauthored work counts once for each author. Material runs through the scan date 2026-09-02.

Every ranked researcher: scientific volume by rank, public footprint in colour
x: rank 1–100 · y: scientific volume (distinct on-topic titles) · bubble size: scholarly pages corroborating the works · hover for detail
Colour = verified non-scientific output pages:0 (none verified in this scan)1234 (maximum)
Scientific volume falls steeply: 18 titles at rank 1, 10–11 by rank 5, and a long tail of 2 from rank 63 to rank 100 — arxiv.org
Only one researcher reaches the maximum of 4 verified public outputs — Jee-Weon Jung, who pairs them with 10 scientific titles — jungjee.com
61 of the 100 have zero verified non-scientific pages in this scan; the 39 with at least one account for all 69 verified output pages — arxiv.org
High corroboration does not track rank: Hemlata Tak (rank 3) has 56 scholarly pages behind 11 titles, more than the leader's 33 — ar5iv.labs.arxiv.org

Method and caveats

What was measured. Scientific volume counts distinct on-topic scholarly work titles attributed to each researcher in the retrieved corpus; scientific mentions counts distinct scholarly source pages corroborating them. Non-scientific volume counts verified, substantive non-journal pages tied to the person: news, interviews and podcasts, talks, blogs, industry guidance, patents, tool or dataset documentation, policy work, and professional profiles. Topic scope covers audio and speech deepfakes, cloned or synthetic voice, spoofed or replayed speech, and voice-fraud or scam-call detection; it excludes visual-only deepfakes and generic speaker recognition or speech synthesis without a detection focus.

How to read the ranking. Ordering is lexicographic — scientific volume, then non-scientific volume, then scholarly mentions — a two-axis ranking, not a weighted composite. It measures retrieved evidence volume, never citations, impact, quality, or absolute productivity.

Limits. This is a reproducible live-web scan, not a canonical census such as OpenAlex or Scopus. Search and extraction can miss work; initials and common names remain ambiguous (an initials-only entry is not expanded); title variants may inflate counts despite normalization. Non-scientific counts are deliberately conservative: a zero means none verified in this scan, not none exists. Scam-audio-specific literature is sparse and the field often writes under adjacent terms such as speech spoofing, synthetic-speech detection, and voice fraud.

The full shortlist

#ResearcherSci. vol.Public vol.Sci. pagesYearsRepresentative publicationRepresentative public outputSources

Live-web evidence ranking, 100 researcher rows plus one aggregate row. Scan through 2026-09-02 over 12 query families: 1,849 unique source pages, 8,370 author–work records, title-level topic filtering. Scientific volume = distinct on-topic scholarly titles per researcher (researcher-level; coauthored works count for each author); non-scientific volume = verified public output pages only. Table shows one representative publication and output per person for space; more exist in the underlying rows.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.

Keenable SELECTAsk your own question
Made with Keenable SELECT