Computer vision researcher working on GPU-efficient instance segmentation and 3D vision-language scene benchmarks
2
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
0
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-21) 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-output data is insufficient for scoring: this snapshot reports no reliable cumulative work counts, venue tiers, or authorship volumes for this name resolution. This axis would measure total publications, their venues, and first/corresponding-author output over time.
Impact—
Citation-based impact cannot be scored: the snapshot carries no usable citation counts, h-index, or i10-index for this researcher. This axis would measure citations, citing works, and uptake of the reviewed papers by other researchers.
Focus100/100
The focus score is 100, the maximum on the 0-100 scale, reflecting a highly concentrated recent record. Both three-year works fall in computer vision, the leading OpenAlex topic “Advanced Neural Network Applications” accounts for 2 tagged works, and the reviewed field distribution points entirely to computer vision (efficient inference / GPU systems), with the remaining topics in adjacent areas.
Collaboration56/100
Collaboration scores 56, with 16 distinct coauthors recorded against only 2 three-year works. Recurring names include Anubhav Khanal, Pramish Paudel, Francois Rameau, Carmelo Scribano, Filippo Muzzini, Nedyalko Prisadnikov and Mohammad Mahdi. Both catalogued 2026 papers are multi-authored, and Paudel is neither first nor corresponding author on either — middle on the GPU decoding paper, last on SceneBench — consistent with a contributing role within multi-author teams.
Activity32/100
Activity scores 32, a comparatively low value that follows directly from the raw counts: only 2 works are recorded in the last three years, and the OpenAlex counts-by-year series lists just 2026 with 2 publications, with no earlier years in this snapshot. Both items are arXiv preprints, one on GPU-accelerated decoding and one on 3D vision-language benchmarking.
Reproducibility55/100
Reproducibility scores 55, computed as a mean over 1 assessed paper. The reviewed work on accelerated decoding of centroid positional encoding carries a medium readiness rating: the reported CUDA rewrite cuts decoding from 196 ms to 31 ms with up to 55% end-to-end speedup at near-equal PQ, but Evolve notes that a 4.5× memory increase and a 16×16-bin PQ drop are left unexplained, tempering replication confidence.
Research summary
Danda Pani Paudel is a computer vision researcher whose recent work concentrates on efficient GPU inference and 3D scene understanding. The Evolve catalog reviews two 2026 arXiv preprints: “Accelerated Decoding of Centroid Positional Encoding for Instance Segmentation”, on which Paudel is a middle author, and “SceneBench: A Hierarchical Benchmark for Vision-Language Understanding of 3D Scenes”, on which Paudel is the last author. OpenAlex topic tags weight “Advanced Neural Network Applications” highest (2 works), with one work each in advanced vision and imaging, image and video retrieval, domain adaptation and few-shot learning, and multimodal machine learning. The snapshot records 16 distinct coauthors, 2 works in the last three years, a focus score of 100, an activity score of 32, a collaboration score of 56, and a reproducibility score of 55 computed over one assessed paper with medium readiness. Output and impact remain unscored because citation statistics are unavailable in this snapshot. This profile is name-resolved from public bibliographic data.
Collaboration pattern
OpenAlex counts 16 distinct coauthors against only 2 recent works, and the collaboration score is 56. Recurring coauthor names include Anubhav Khanal, Pramish Paudel, Francois Rameau, Carmelo Scribano, Filippo Muzzini, Nedyalko Prisadnikov and Mohammad Mahdi. On both catalogued 2026 papers Paudel is neither first nor corresponding author — middle author on the GPU decoding paper and last author on SceneBench — a pattern consistent with contributing within multi-author teams rather than leading them.
Research keywords
Highlights
CUDA rewrite cuts decoding from 196 ms to 31 ms
Unexplained memory and accuracy trade-offs flagged
Last author on SceneBench, a hierarchical 3D scene benchmark
In 2026, Paudel appears as last (non-corresponding) author of the arXiv preprint “SceneBench: A Hierarchical Benchmark for Vision-Language Understanding of 3D Scenes”; Evolve has not yet published a review verdict for this paper.
Timeline
- 2026arXiv preprint “Accelerated Decoding of Centroid Positional Encoding for Instance Segmentation” appears, with Paudel as a middle author.
- 2026arXiv preprint “SceneBench: A Hierarchical Benchmark for Vision-Language Understanding of 3D Scenes” appears, with Paudel as last author.
- 2026OpenAlex records 2 works in 2026 — the only year present in this snapshot, both in computer vision.
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 derive from OpenAlex records only; no other bibliographic sources were used.
- Output and impact are unscored because citation counts, h-index and i10-index are unavailable for this name resolution in this snapshot.
- Coverage is limited to two 2026 arXiv preprints; no peer-reviewed venue, institution, country, or ORCID data is present.
- Identity is resolved from public bibliographic data only, so disambiguation from same-name researchers is not guaranteed.