Thai low-resource language systems researcher spanning streaming ASR, synthetic-data OCR, and fixed-voice TTS
3
papers here
3
reviewed by Evolve
2
first-author
1
corresponding
—
works · OpenAlex
2
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-16) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
Publication-volume data for this axis is not yet available. Output would measure the quantity and consistency of the researcher's papers over time, including venue-normalized productivity; Evolve has not computed it and does not estimate a score here.
Impact—
Citation-impact data is insufficient for scoring. Impact would measure citation-based influence — field-weighted citation counts, percentile placement, and uptake by other researchers — none of which are available beyond h-index 2 and i10-index 0; no score is estimated.
Focus0/100
The focus score of 0 reflects an unusually even spread: the field distribution assigns exactly 0.333 each to speech processing/ASR, document AI and OCR, and text-to-speech. Topic counts reinforce the breadth — topic modeling (21), NLP techniques (17), speech recognition and synthesis (14), multimodal ML (12), dialogue systems (11) — spanning several Thai-language modalities rather than one specialty.
Collaboration41/100
The collaboration score is 41, built on a compact network of 7 distinct coauthors over the last three years. Recurring partners include Warit Sirichotedumrong, Sittipong Sripaisarnmongkol, Tanawin Samutsin, Shah Faisal Wani, Potsawee Manakul, Pakorn Nathong, and Phatrasek Jirabovonvisut. All three reviewed papers are multi-authored, with the researcher moving between first, corresponding, and last-author positions.
Activity100/100
Activity scores 100, the maximum, driven by 36 works in the last three years. Yearly counts show a sharp acceleration: 1 in 2016, 1 in 2023, 2 in 2024, then 19 in 2025 and 15 already in 2026 — a roughly tenfold jump from 2024 to 2025 that marks an intensive current publishing phase.
Reproducibility70/100
Reproducibility averages 70 across three Evolve-assessed papers. Typhoon ASR Streaming is rated high readiness, though gains are single-set and modest; the synthetic-data Thai OCR and fixed-voice Thai TTS papers are rated medium — careful ablations, but missing significance tests or perceptual evaluation. The pattern indicates solid artifact-grounded work with evaluation-methodology gaps.
Research summary
Kunat Pipatanakul is a researcher focused on Thai-language technologies, whose OpenAlex-linked bibliometric record splits evenly across three fields: speech processing/automatic speech recognition, computer vision/NLP for document AI and OCR, and text-to-speech, each at 0.333 of the field distribution. The profile shows a steep recent ramp: 36 works in the last three years, with annual counts moving from 1 (2016) and 1 (2023) to 2 (2024), 19 (2025), and 15 (2026) — the basis for a maximum activity score. Citation indicators remain early-stage (h-index 2, i10-index 0), so output and impact are left unscored rather than estimated. Evolve has reviewed three 2026 arXiv papers. Typhoon ASR Streaming (last author) is judged deployable: cache-aware models cut streamed CER 4.3-4.5x and n-gram fusion lifts keyword recall ~4 points, on a single evaluation set. How Far Can Synthetic Data Take Thai OCR? (first and corresponding author) claims synthetic-only training yields competitive Thai OCR, supported by careful ablations but machine-generated references and no significance tests. Building and Evaluating Fixed-Voice Thai TTS from Synthetic Speech (first author) distills a 15s reference into an 82M fixed-voice model. Mean reproducibility across the three is 70.
Collaboration pattern
Collaboration runs through a small, stable Thai-language NLP circle: 7 distinct coauthors over the last three years, including Warit Sirichotedumrong, Sittipong Sripaisarnmongkol, Tanawin Samutsin, Shah Faisal Wani, Potsawee Manakul, Pakorn Nathong, and Phatrasek Jirabovonvisut. Within the reviewed papers the researcher leads hands-on as first author twice and corresponding author once, while taking the last-author slot on the streaming ASR systems work — a mix of direct authorship leadership and senior placement on systems releases.
Research keywords
Highlights
Deployable streaming Thai ASR with cache-aware shallow fusion
Synthetic-only training makes Thai OCR competitive
Fixed-voice Thai TTS distilled from synthetic speech
Timeline
- 2026First-authored 'Building and Evaluating Fixed-Voice Thai TTS from Synthetic Speech', distilling a 15s reference into an 82M-parameter fixed-voice model.
- 2026First-authored and corresponding-authored 'How Far Can Synthetic Data Take Thai OCR?', arguing synthetic-only data suffices for competitive Thai OCR.
- 2026Last-authored 'Typhoon ASR Streaming', a steerable low-latency Thai speech recognition system with real-time shallow fusion.
- 2026Concentrated all three Evolve-reviewed arXiv releases in a single year across ASR, OCR, and TTS.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Bibliometrics come from an OpenAlex snapshot only; h-index (2) and i10-index (0) reflect that coverage, not total career output.
- Output and impact dimensions are unscored because the underlying data is insufficient.
- Institution, country, and ORCID were not provided; this profile is name-resolved from public bibliographic data, and the name may be shared by others.
- Highlights rely on the three Evolve-reviewed 2026 arXiv papers; other works within the 36 recent outputs are not individually assessed.