Generative audio mixing and audio-visual segmentation; first author on 2026 arXiv preprints, Tampere UAS, Finland
2
papers here
0
reviewed by Evolve
2
first-author
0
corresponding
—
works · OpenAlex
1
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-28) 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—
Not yet scored: this axis aggregates total publication volume and output growth. OpenAlex lists only a handful of indexed works (h-index 1, four works in the last three years), too thin a record for a reliable output score.
Impact—
Not yet scored: this axis measures citation-based influence, such as citation counts and h-index trajectories. With an h-index of 1 and an i10-index of 1, the citation record is too sparse to support a meaningful impact score.
Focus—
Not yet scored: this axis tracks topical concentration versus dispersion across a researcher's works. Five OpenAlex topic labels are present, but with so few indexed works the focus score cannot yet be computed reliably.
Collaboration41/100
The radar collaboration score is 41, derived from seven distinct co-authors in the record. Both cataloged 2026 preprints list Viertola as first author, non-corresponding, indicating a team contributor role rather than sole senior authorship. Frequent co-authors include Giulio Cengarle, Gouthaman KV, Daniel Arteaga, Lie Lu, Vladimir Iashin, Sophie Tötterström, and Esa Rahtu, a mix spanning the audio and audio-visual projects; the moderate score reflects a small but active collaborative network.
Activity47/100
The radar activity score is 47. OpenAlex records four works in the last three years, with per-year counts of one in 2021, 2024, and 2025 and two in 2026; both 2026 items are first-author arXiv preprints in the Evolve catalog. The trajectory suggests accelerating output in 2026 after a sparse earlier record, though overall volume remains small and the score sits in the middle range.
Reproducibility—
Not yet scored: this axis evaluates artifact availability and reproducibility signals. Evolve's per-paper reviews flag one 2026 preprint as high reproducibility readiness and the other as low, but a profile-level axis score is not available.
Research summary
Ilpo Viertola is a Finland-based researcher affiliated with Tampere University of Applied Sciences whose public bibliometric record is early-stage but clearly oriented toward machine learning for audio and vision. The OpenAlex identity record shows an h-index of 1, an i10-index of 1, four works in the last three years, and a sparse per-year footprint (one work each in 2021, 2024, and 2025; two indexed for 2026), alongside seven distinct co-authors. Topic clustering is led by image and signal denoising methods, generative adversarial networks and image synthesis, music and audio processing, and music technology and sound studies, with one entry in service-oriented architecture and web services. The Evolve catalog contains two 2026 arXiv preprints on which Viertola is first author. 'Spot, Separate, and Enhance' presents a fully generative audio-mixing approach built on conditional flow matching and a diffusion transformer; Evolve's review reports that it beats VisAH on generative metrics and preference while trailing VisAH-FM on reconstruction metrics. 'Less is More' proposes EASE, an encoder-only audio-visual segmentation model achieving state-of-the-art AVSS at three times the throughput, and is flagged high reproducibility readiness versus low for the first paper. Radar scores are moderate for activity (47) and collaboration (41); the output, impact, focus, and reproducibility axes lack sufficient data to score.
Collaboration pattern
Seven distinct co-authors appear in the record, supporting a radar collaboration score of 41. Frequent co-authors span both research threads: Giulio Cengarle, Gouthaman KV, Daniel Arteaga, Lie Lu, Vladimir Iashin, Sophie Tötterström, and Esa Rahtu. Both cataloged 2026 preprints list Viertola as first author but not corresponding author, indicating a hands-on contributor role within larger teams — one tied in Evolve's review to Dolby's SSE work, the other to encoder-only audio-visual segmentation.
Research keywords
Highlights
EASE: encoder-only audio-visual segmentation with SotA AVSS at 3x throughput
Fully generative audio mixing with conditional flow matching
First-author streak across two 2026 preprints
Both 2026 arXiv preprints in the Evolve catalog list Viertola as first author (non-corresponding), one in generative audio mixing and one in audio-visual segmentation — a concentrated burst matching the record's two works indexed for 2026.
Timeline
- 2021Earliest indexed activity in the OpenAlex record: one work counted for 2021.
- 2024One work indexed for 2024, indicating the record remained sparse through the mid-2020s.
- 2025One work indexed for 2025, maintaining a low-volume cadence before the 2026 step-up.
- 2026Busiest indexed year with two works, both first-author arXiv preprints in the Evolve catalog: the fully generative audio-mixing paper and the encoder-only audio-visual segmentation model EASE.
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.
- Bibliometric scores and raw counts (h-index, i10-index, per-year counts, topics) come from an OpenAlex identity record; other indexes may differ.
- Only the two 2026 preprints carry Evolve review verdicts; other works in the record are not individually assessed.
- Output, impact, focus, and reproducibility radar axes are uncomputed owing to insufficient underlying data.
- Identity is name-resolved from public bibliographic data (ORCID 0009-0006-2361-8067); no biographical information beyond the provided record is used.