Nick Kashani Motlagh
Publications

Four peer-reviewed first-author papers.

Peer-reviewed

Published papers

Machine Vision and Applications / 2025

First author

Naturally Constrained Reject Option Classification

N. Kashani Motlagh, J. Davis, T. Anderson, J. Gwinnup

Journal extension evaluating per-class binomial reject thresholds across 4 vision, 3 text and 8 synthetic datasets, from 2 to 1,000 classes.

  • vision
  • reject option
  • calibration
  • ImageNet
  • Extends the natural reject-region constraint across controlled synthetic data, benchmark image classification, and text classification.
  • Reports transfer behavior on CINIC10 and long-tailed iNaturalist19 alongside explicit selective-accuracy, reject-accuracy, and coverage tradeoffs.
Paper arXiv Code
Related code and data

learning-idk

WMT 2024 / 2024

First author

Assessing the Role of Imagery in Multimodal Machine Translation

N. Kashani Motlagh, J. Davis, J. Gwinnup, G. Erdmann, T. Anderson

The best multimodal translation system reads the image correctly 81% of the time, but that only changes its translation 63% of the time. 'Images don't help' was partly a measurement artifact.

  • multimodal MT
  • vision-language models
  • evaluation
  • WMT
  • Introduced imagery-aware contrastive probes for testing whether model scores change under mismatched visual context.
  • Evaluated three English-to-French multimodal model families, plus gated variants, under matched and mismatched visual context.
Paper

ISVC 2022 / 2022

First author Springer Best Paper Award

Learning When to Say “I Don’t Know”

N. Kashani Motlagh, J. Davis, T. Anderson, J. Gwinnup

Per-class abstention thresholds that need no rejection cost or coverage target: CIFAR-100 selective accuracy climbs from 88.3% to 97.8% at 77.3% coverage.

  • vision
  • reject option
  • selective accuracy
  • ImageNet
  • Springer Best Paper Award at ISVC 2022; later extended in the MVA 2025 journal version.
  • At the ImageNet δ=.75 operating point, B-CDF improved select accuracy by 0.4 percentage points and coverage by 1.3 points versus an uncalibrated global 0.5 threshold.
Paper arXiv Code
Related code and data

learning-idk

ICCV 2021 Workshop on LUAI / 2021

First author

A Framework for Semi-automatic Collection of Temporal Satellite Imagery for Analysis of Dynamic Regions

N. Kashani Motlagh, A. Radhakrishnan, J. Davis, R. Ilin

OpenStreetMap-guided imagery collection and labeling tools for building temporal satellite datasets for dynamic-region analysis.

  • remote sensing
  • data collection
  • labeling pipelines
  • change detection
  • Combined OpenStreetMap extraction, polygon filtering, and temporal imagery download scripts in a public Python workflow.
  • Released the collection code and a sample construction-site satellite-imagery dataset under the repository's GPL license.
Paper Code
Related code and data

construction-site-satellite-imagery-collection

Not yet peer-reviewed

Current manuscripts

Dissertation chapter · arXiv preprint in preparation / 2026

First author

When Retrieval Makes the Answer Worse

Nick Kashani Motlagh

Retrieval-based refinement repairs 10.8% of draft answers and destroys 8.1% that were already correct. Confidence cannot tell you which you are about to get.

  • selective prediction
  • adaptive QA
  • retrieval-augmented generation
  • abstention
  • Across 25,870 held-out questions on NQ-Open, TriviaQA and PopQA, retrieval-based refinement repairs 10.8% of answers and harms 8.1% — every dataset and retriever slice contains both.
  • Confidence does not predict recoverability: a verifier that sorts draft accuracy from 4.4% to 93.9% across deciles still values refinement at +23.7 points in the lowest decile and net negative in the top four.