Pith. sign in

REVIEW 4 major objections 5 minor 62 references

Enhanced Dermatology Image Quality Assessment via Cross-Domain Training

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that training an image-quality model jointly on natural and dermatology image datasets outperforms dermatology-only training for predicting the quality of skin images, and supports this with a new 1,800-image dermatology…

desk verdict The paper's own tables undercut its central claim about cross-domain training, but the new dermatology IQA dataset and systematic evaluation still deserve peer review with major revisions. read the letter →

arxiv 2506.16116 v1 pith:RYKDRTUQ submitted 2025-06-19 eess.IV cs.CVcs.MM

Ignacio Hernández Montilla , Alfonso Medela , Paola Pasquali , Andy Aguilar , Taig Mac Carthy , Gerardo Fernández , Antonio Martorell , Enrique Onieva This is my paper · ORCID
classification eess.IVcs.CVcs.MM
keywords imagequalityassessmentteledermatologycross-domaintrainingno-referencedeeplearningmeanopinionscoredermatologyimagingEfficientNet
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Teledermatology works only when patient photos are good enough for a clinician to read, but automated quality checking for skin images has little data and has not used the large natural-image quality databases that exist for everyday photos. This paper tries to establish that a model trained jointly on natural image-quality datasets and a new dermatology image-quality dataset outperforms a model trained on dermatology data alone, while staying accurate on natural images. To make the test possible, the authors built and rated a new dermatology benchmark from 100 pristine skin images distorted in seven ways, producing 1,800 annotated images with 40 human ratings each. If the claim holds, teledermatology platforms could automatically flag poor-quality patient photos before review, and other small medical-imaging domains could borrow the same cross-domain recipe.

What carries the argument

The mechanism is cross-domain training of a no-reference IQA regression model: a convolutional EfficientNet (B0, B3, or B5) with two fully-connected layers, trained with mean squared error loss to predict a scalar quality score, and evaluated with Pearson's linear correlation coefficient and Spearman's rank correlation coefficient. The enabling data object is the new Legit.Health-DIQA-Artificial dataset, built by taking 100 pristine dermatology images and applying seven distortion families (JPEG compression, Gaussian blur, pixelation, sharpening, brightness, color, contrast) at multiple strengths, yielding 1,800 images annotated by 40 non-expert observers on a 1-to-10 scale; subject-based splits prevent leakage between training and test views. To combine databases, the paper linearly rescales each natural dataset's scores to the 1-to-10 range and reweights classes to correct score imbalance. The argument turns on comparing single-domain models, natural-only models, and 'All' models across the same test sets.

What would settle it

Re-run the cross-domain versus dermatology-only comparison after replacing naive linear rescaling of mean opinion scores with dataset-specific calibration, for instance a per-dataset monotonic mapping to a common perceptual scale; if the 'All' models then no longer outperform dermatology-only models on Legit.Health-DIQA-Artificial, the reported benefit is an artifact of score rescaling rather than genuine cross-domain transfer.

Watch

Extended reading notes

Core claim

The central claim is that cross-domain training, training one no-reference image quality model simultaneously on several natural image quality databases and a new dermatology image quality database, yields better dermatology image quality prediction than training on dermatology data alone, while preserving or improving performance on natural images. The authors support this with a new benchmark, Legit.Health-DIQA-Artificial: 100 pristine dermatology images from multiple sources, each distorted in seven ways at multiple strengths to produce 1,800 rated images, with 40 non-expert ratings per image averaged into a mean opinion score. Across five random splits, EfficientNet models trained on all datasets ('All') achieve the highest average correlation across test sets and outperform single-domain models on the dermatology test set in most configurations, although the largest EfficientNet-B5 variants are noted as the stated exception.

Load-bearing premise

The load-bearing premise is that linearly rescaling each natural image-quality database's mean opinion scores to the same 1-to-10 range makes scores from different databases comparable enough that pooling them into one training set helps rather than hurts; the paper itself concedes that without dataset-specific scaling, images with different quality features can end up with similar scores.

Editorial extensions

If this is right

  • In most model sizes tested, the 'All' cross-domain models beat both natural-only and dermatology-only models on the new dermatology test set; the largest EfficientNet-B5 variant is the stated exception, where the paper suspects overfitting.
  • No single natural dataset is enough: the best natural-only model (SPAQ) reaches roughly 0.64 correlation on dermatology images, well below the cross-domain models, so relying on a natural IQA model alone is not reliable for teledermatology.
  • Cross-domain models also keep strong accuracy on natural test sets, roughly 0.86 to 0.96 PLCC on KonIQ-10k, SPAQ, and GFIQA-20k, so one model can serve multiple domains instead of requiring a separate dermatology model.
  • Increasing input resolution and model size helps cross-domain training on most datasets, whereas for single-domain training the effect is dataset-dependent.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, the scaling question is left open; a direct next experiment would replace naive linear rescaling with dataset-specific MOS calibration and check whether the cross-domain advantage on dermatology images grows, shrinks, or disappears.
  • Because the new dermatology dataset contains only artificial distortions, the practical benefit for teledermatology still needs confirmation on authentic patient-taken photos with real blur, lighting, and framing problems.
  • The poor mutual transfer between the two artificial-distortion datasets, Kadid-10k and the new DIQA set, hints that single-domain models may be memorizing content rather than distortions; this could be tested by measuring performance on distorted versions of unseen content.
  • If the cross-domain recipe generalizes, other small medical imaging fields with scarce quality ratings, such as retinal or dermoscopic imaging, could follow the same pattern: pool large natural IQA databases with a small curated medical set rather than collecting thousands of medical ratings.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. This paper proposes to improve no-reference image quality assessment (IQA) for dermatology by jointly training an EfficientNet on natural-domain IQA databases (KonIQ-10k, SPAQ, GFIQA-20k, Kadid-10k, BIQ2021, LIVE-ItW) and on a newly introduced artificially distorted dermatology image database, Legit.Health-DIQA-Artificial, annotated by 40 non-expert observers. The authors compare three model sizes (B0, B3, B5) trained on each single dataset and on all datasets combined, reporting PLCC and SROCC on every test set across five random splits. They conclude that cross-domain training yields optimal performance across domains and improves dermatology IQA relative to dermatology-only training.

Significance. The new dermatology IQA dataset, despite being limited to artificial distortions, fills a gap in the field and could serve as a benchmark for future work. The use of subject-wise splits to prevent leakage between a pristine image and its distorted views is a methodological strength that improves the reliability of the internal comparisons. However, the central claim of the paper is not supported by its own results: B5 trained only on the dermatology dataset outperforms B5 trained on all datasets, and single-domain models outperform the All model on some natural datasets. The lack of significance testing and the acknowledged MOS rescaling problem further weaken the evidence. If the claims were properly scoped and supported by statistical analysis, this could be a useful contribution to teledermatology and cross-domain IQA.

major comments (4)
  1. [Tables 3 and 4; Section 5] The central claim that cross-domain training 'provides better results than training exclusively on dermatology data' is contradicted by the EfficientNet-B5 results on the DIQA test set: B5 trained only on DIQA reaches PLCC 0.8666±0.0400 and SROCC 0.8466±0.0420, whereas B5 trained on All reaches PLCC 0.8363±0.0349 and SROCC 0.8125±0.0364. The paper's dismissal of this as 'overfitting' (Section 5) is unexplained; the DIQA-only B5 actually shows the highest in-domain correlation, so additional natural data appears to hurt rather than help at this capacity. The conclusion must either be restricted to B0/B3 or supported by an analysis of why the B5 All model underperforms.
  2. [Table 3; Abstract] The abstract's claim that cross-domain training 'yields optimal performance across domains' is not supported by the paper's own tables: single-domain models outperform All on their own test sets in several cases, e.g., B0 KonIQ-10k PLCC 0.9134 vs. 0.8759, B0 SPAQ 0.9133 vs. 0.9052, B0 GFIQA-20k 0.9648 vs. 0.9610, and B0 KonIQ-10k SROCC 0.8925 vs. 0.8427. If the claim means something weaker, such as 'competitive on natural datasets and improved on the dermatology dataset for smaller models,' it should be stated precisely.
  3. [Section 3.3; Section 5] The validity of the cross-domain training rests on the commensurability of the MOS scores from different datasets after linear rescaling to [1,10]. The paper itself concedes in the Discussion that without dataset-specific MOS scaling, images from different quality-feature distributions may be assigned similar scores, reducing the effectiveness of combining datasets. Since this admission directly applies to the central experiment, the paper should either demonstrate that the reported improvements are robust to alternative rescaling schemes, or temper the conclusion accordingly.
  4. [Section 4] No statistical significance testing is reported: all conclusions are drawn from means and standard deviations over five random splits, and some differences central to the claims fall within one standard deviation or have overlapping intervals (e.g., B0 All vs. B0 DIQA on DIQA PLCC 0.8021±0.0550 vs. 0.7470±0.0525). The paper should report paired significance tests (e.g., Wilcoxon signed-rank or paired t-test across splits) for the comparisons that underlie the main claim.
minor comments (5)
  1. [Discussion and Section 3.3] Typographical errors: 'entirely different form a natural image' should be 'entirely different from a natural image', and 'independent form each other' should be 'independent from each other'.
  2. [Conclusion] In the concluding section, 'DIQA modles' should be corrected to 'DIQA models'.
  3. [Data Availability Statement] The statement that the dataset and code are not publicly available limits reproducibility; the authors should at least release the quality ratings in anonymized form or explain the intellectual property restrictions in more detail.
  4. [Section 3.1 and Table 1] The paper does not compare against any existing NR-IQA baselines (e.g., BRISQUE, NIQE, or other deep learning models), so the absolute correlation values are difficult to interpret; adding such baselines would strengthen the evaluation.
  5. [Section 3.3] The description of the splitting strategy is confusing: Kadid-10k is said to have image clusters, yet the next sentence states that 'all images from Kadid-10k, KonIQ-10k, GFIQA-20k, and SPAQ are independent samples'; please clarify whether the split unit is a cluster or an individual image for each dataset.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the cross-domain comparison is an empirical benchmark with subject-wise splits; self-citation [32] is motivational, not load-bearing.

full rationale

The paper's central claim (cross-domain training improves dermatology IQA) is evaluated with held-out test splits, not by construction. Section 3.3 explicitly describes subject-wise stratification: 'if a pristine image is in the training set, all its distorted views will also be put in that set', preventing label leakage between DIQA training and DIQA test sets. Tables 3 and 4 compare models trained under the same protocol (EfficientNet B0/B3/B5, AdamW, MSE loss), and no reported number is a re-statement of a fitted parameter or of the MOS rescaling in Section 3.3. The linear rescaling of natural MOS to [1,10] is a data-preprocessing choice that the authors themselves flag as a limitation ('Without proper and dataset-specific scaling... can reduce the effectiveness of combining IQA datasets'), not a fitted quantity later presented as a prediction. Reference [32] is the authors' preliminary work, but it is cited only as background and motivation; the current comparison does not rely on [32] for the new empirical result. The B5 reversal (DIQA-only PLCC 0.8666 vs All 0.8363 on DIQA) is an internal-consistency problem for the paper's conclusion, but it is not circular: the outcome was not forced by the setup, and the fact that some configurations contradict the claim shows the comparison is genuine. No equation is defined in terms of a target result, no fitted value is renamed as a prediction, and no uniqueness theorem is imported from the authors' prior work. Therefore the derivation chain is self-contained and no significant circularity is present.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The central claim rests on empirical training rather than derivation, so the ledger records the dataset construction choices, the scaling assumption, and the hyperparameters that are load-bearing for the cross-domain comparison.

free parameters (8)
  • learning_rate = 2e-4
    Chosen via learning rate tests (Section 3.4); used in all experiments.
  • weight_decay = 1e-5
    AdamW weight decay set in Section 3.4.
  • training_epochs = 20
    Fixed training budget in Section 3.4.
  • dropout_rate = 0.50
    Dropout in the fully connected layers, Section 3.4.
  • augmentation_resize_margin = 12.5% larger than input
    Initial resizing for augmentation, Section 3.4.
  • pristine_image_count = 100
    Chosen selection size for the new dataset, Section 3.2.1.
  • observer_count = 40
    Board size for MOS annotation, Section 3.2.2.
  • distortion_level_counts = JPEG 3, blur 1, pixelation 2, sharpening 3, brightness 4, color 2, contrast 3
    Manually chosen distortion strengths, Section 3.2.1 and Figure 1.
assumptions (5)
  • domain assumption Linear rescaling of MOS to [1,10] makes scores across datasets commensurable for pooled training.
    Section 3.3 applies rescaling; the Discussion admits dataset-specific MOS scaling is needed but omitted, so the validity of pooling is unverified.
  • domain assumption Non-expert observers can give valid perceptual quality scores for dermatology images without dermatology expertise.
    Section 3.2.2 states quality features exist beyond semantic content; Section 5 speculates that aesthetics and clinical utility may affect ratings, which weakens this assumption.
  • domain assumption Artificial distortions from Kadid-10k (JPEG, blur, pixelation, sharpening, brightness, color, contrast) represent distortions relevant to real teledermatology photos.
    Section 3.2.1; Section 6 says authentic-distortion datasets are future work, so this premise is not validated.
  • domain assumption ImageNet-pretrained EfficientNet features transfer to image quality prediction.
    Section 3.4 initializes all models from ImageNet; no ablation without pretraining is given.
  • domain assumption Random horizontal flipping and small resizing/cropping do not break the image-to-MOS relationship.
    Section 3.4 follows NIMA; no verification is provided for dermatology images.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhanced Dermatology Image Quality Assessment via Cross-Domain Training." pith.science (2026). https://pith.science/paper/RYKDRTUQ

@misc{pith2026250616116,
  author       = {Pith},
  title        = {Pith review of: Enhanced Dermatology Image Quality Assessment via Cross-Domain Training},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RYKDRTUQ}},
  note         = {Machine review of arXiv:2506.16116}
}
read the original abstract

Teledermatology has become a widely accepted communication method in daily clinical practice, enabling remote care while showing strong agreement with in-person visits. Poor image quality remains an unsolved problem in teledermatology and is a major concern to practitioners, as bad-quality images reduce the usefulness of the remote consultation process. However, research on Image Quality Assessment (IQA) in dermatology is sparse, and does not leverage the latest advances in non-dermatology IQA, such as using larger image databases with ratings from large groups of human observers. In this work, we propose cross-domain training of IQA models, combining dermatology and non-dermatology IQA datasets. For this purpose, we created a novel dermatology IQA database, Legit.Health-DIQA-Artificial, using dermatology images from several sources and having them annotated by a group of human observers. We demonstrate that cross-domain training yields optimal performance across domains and overcomes one of the biggest limitations in dermatology IQA, which is the small scale of data, and leads to models trained on a larger pool of image distortions, resulting in a better management of image quality in the teledermatology process.

Figures

Figures reproduced from arXiv: 2506.16116 by the authors.

Figure 2
Figure 2. Although this may be considered a medical image dataset, per￾ceptual image quality assessment should not require prior expert knowledge in dermatology, as quality features exist beyond the semantic content of an image. Although perceptual visual quality can be content dependent [42], we believe that common distortions such as overexposure, underexposure, blur, or compression can be easily detected by any observer in… view at source ↗
Figure 1
Figure 1. A pristine image (top-left image) of the Legit.Health [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Distribution of Mean Opinion Scores (MOS) for each natural IQA dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Analysis of the impact of input image size on corre [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 56 canonical work pages

  1. [1]

    Nisar Ahmed and Shahzad Asif. 2022. BIQ2021: a large-scale blind image quality assessment database. Journal of Electronic Imaging 31, 5 (2022), 053010–053010

    work page 2022

  2. [2]

    Marine Amouroux, Sébastien Le Cunff, Alexandre Haudrechy, and Walter Blondel

    work page

  3. [3]

    Autrusseau, Florent, and Stütz, Thomas, and Pankajakshan, Vinod. 2010. Subjec- tive quality assessment of selective encryption techniques. http://www.polytech. univ-nantes.fr/autrusseau-f/Databases/SelectiveEncryption/s/

    work page 2010

  4. [4]

    Igor Barros Barbosa, Theoharis Theoharis, Christian Schellewald, and Cham Ath- wal. 2013. Transient biometrics using finger nails. In2013 IEEE Sixth International Conference on Biometrics: Theory, Applications and Systems (BTAS) . IEEE, 1–6

    work page 2013

  5. [5]

    Igor Barros Barbosa, Theoharis Theoharis, and Ali E Abdallah. 2016. On the use of fingernail images as transient biometric identifiers: Biometric recognition using fingernail images. Machine Vision and Applications 27 (2016), 65–76

    work page 2016

  6. [6]

    Simone Bianco, Luigi Celona, Paolo Napoletano, and Raimondo Schettini. 2018. On the use of deep learning for blind image quality assessment. Signal, Image and Video Processing 12 (2018), 355–362

    work page 2018

  7. [7]

    Mollie R Cummins, Triton Ong, Julia Ivanova, Janelle F Barrera, Hattie Wilczewski, Hiral Soni, Brandon M Welch, and Brian E Bunnell. 2023. Con- sensus Guidelines for Teledermatology: Scoping Review. JMIR dermatology 6 (2023), e46121

    work page 2023

  8. [8]

    J Dahlén Gyllencreutz, E Johansson Backman, K Terstappen, and J Paoli. 2018. Teledermoscopy images acquired in primary health care and hospital settings– a comparative study of image quality. Journal of the European Academy of Dermatology and Venereology 32, 6 (2018), 1038–1043

    work page 2018

Show all 62 references
  1. [9]

    Kanjar De and V Masilamani. 2013. A new no-reference image quality measure for blurred images in spatial domain. Journal of Image and Graphics 1, 1 (2013), 39–42

    2013

  2. [10]

    Yuming Fang, Hanwei Zhu, Yan Zeng, Kede Ma, and Zhou Wang. 2020. Perceptual quality assessment of smartphone photography. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 3677–3686

    2020

  3. [11]

    Deepti Ghadiyaram and Alan C Bovik. 2015. Massive online crowdsourced study of subjective and objective picture quality. IEEE Transactions on Image Processing 25, 1 (2015), 372–387

    2015

  4. [12]

    Mara Giavina Bianchi, Andre Santos, and Eduardo Cordioli. 2021. Dermatologists’ perceptions on the utility and limitations of teledermatology after examining 55,000 lesions. Journal of telemedicine and telecare 27, 3 (2021), 166–173

    2021

  5. [13]

    Gonzalez, Esther and Alvarez, Luis and Mazorra, Luis. 2024. AMI Ear Database. https://webctim.ulpgc.es/research_works/ami_ear_database/. Online; accessed 01-September-2024

    2024

  6. [14]

    Ignacio Hernández Montilla, Alfonso Medela, Taig Mac Carthy, Andy Aguilar, Pedro Gómez Tejerina, Alejandro Vilas Sueiro, Ana María González Pérez, Laura Vergara de la Campa, Loreto Luna Bastante, Rubén García Castro, et al . 2023. Automatic International Hidradenitis Suppurati...

    2023

  7. [15]

    Daniel T Hogarty, John C Su, Kevin Phan, Mohamed Attia, Mohammed Hossny, Saeid Nahavandi, Patricia Lenane, Fergal J Moloney, and Anousha Yazdabadi

  8. [16]

    Vlad Hosu, Hanhe Lin, Tamas Sziranyi, and Dietmar Saupe. 2020. KonIQ-10k: An ecologically valid database for deep learning of blind image quality assessment. IEEE Transactions on Image Processing 29 (2020), 4041–4056

    2020

  9. [17]

    Esmee Irvine, Leela Sayed, Nick Johnson, and Joseph Dias. 2023. The ability of patients to provide standardized, patient-taken photographs for the remote assessment of Dupuytren disease. HAND 18, 1 (2023), 139–144

    2023

  10. [18]

    Raluca Jalaboi, Ole Winther, and Alfiia Galimzianova. 2023. Explainable image quality assessments in teledermatological photography.Telemedicine and e-Health (2023)

    2023

  11. [19]

    Eric C Larson and Damon M Chandler. 2010. Most apparent distortion: full- reference image quality assessment and the role of strategy. Journal of electronic imaging 19, 1 (2010), 011006–011006

    2010

  12. [20]

    Jonathan J Lee and Joseph C English. 2018. Teledermatology: a review and update. American journal of clinical dermatology 19 (2018), 253–260

    2018

  13. [21]

    Hanhe Lin, Vlad Hosu, and Dietmar Saupe. 2019. KADID-10k: A large-scale artificially distorted IQA database. In 2019 Eleventh International Conference on Quality of Multimedia Experience (QoMEX) . IEEE, 1–3

    2019

  14. [22]

    Hanhe Lin, Vlad Hosu, and Dietmar Saupe. 2020. DeepFL-IQA: Weak supervision for deep IQA feature learning. arXiv preprint arXiv:2001.08113 (2020)

    2020 arXiv

  15. [23]

    Xinwei Liu, Marius Pedersen, and Jon Yngve Hardeberg. 2014. CID: IQ–a new image quality database. In Image and Signal Processing: 6th International Confer- ence, ICISP 2014, Cherbourg, France, June 30–July 2, 2014. Proceedings 6 . Springer, 193–202

    2014

  16. [24]

    Teresa Lopez de Coca, Lucrecia Moreno, Mónica Alacreu, and Maria Sebastian- Morello. 2022. Bridging the generational digital divide in the healthcare environ- ment. Journal of Personalized Medicine 12, 8 (2022), 1214

    2022

  17. [25]

    Remedios López-Liria, María Ángeles Valverde-Martínez, Antonio López-Villegas, Rafael Jesús Bautista-Mesa, Francisco Antonio Vega-Ramírez, Salvador Peiró, and Cesar Leal-Costa. 2022. Teledermatology versus face-to-face dermatology: An analysis of cost-effectiveness from eight ...

    2022

  18. [26]

    Ilya Loshchilov, Frank Hutter, et al. 2017. Fixing weight decay regularization in adam. arXiv preprint arXiv:1711.05101 5 (2017)

    2017 arXiv

  19. [27]

    Taig Mac Carthy, Ignacio Hernández Montilla, Andy Aguilar, Rubén García Castro, Ana María González Pérez, Alejandro Vilas Sueiro, Laura Vergara de la Campa, Fernando Alfageme, and Alfonso Medela. 2024. Automatic Urticaria Activity Score: Deep Learning–Based Automatic Hive Coun...

    2024

  20. [28]

    Kate Maier, Luiz Zaniolo, and Oge Marques. 2022. Image quality issues in teled- ermatology: A comparative analysis of artificial intelligence solutions. Journal of the American Academy of Dermatology 87, 1 (2022), 240–242

    2022

  21. [29]

    Alfonso Medela, Taig Mac Carthy, S Andy Aguilar Robles, Carlos M Chiesa- Estomba, and Ramon Grimalt. 2022. Automatic SCOring of atopic dermatitis using deep learning: a pilot study. JID Innovations 2, 3 (2022), 100107

    2022

  22. [30]

    Teresa Mendonça, Pedro M Ferreira, Jorge S Marques, André RS Marcal, and Jorge Rozeira. 2013. PH 2-A dermoscopic image database for research and benchmarking. In 2013 35th annual international conference of the IEEE engineering in medicine and biology society (EMBC) . IEEE, 5437–5440

    2013

  23. [31]

    Anish Mittal, Anush Krishna Moorthy, and Alan Conrad Bovik. 2012. No- reference image quality assessment in the spatial domain. IEEE Transactions on image processing 21, 12 (2012), 4695–4708

    2012

  24. [32]

    Ignacio Hernández Montilla, Taig Mac Carthy, Andy Aguilar, and Alfonso Medela

  25. [33]

    Krzysztof Okarma. 2019. Current trends and advances in image quality assess- ment. Elektronika ir Elektrotechnika 25, 3 (2019), 77–84

    2019

  26. [34]

    Andre GC Pacheco, Gustavo R Lima, Amanda S Salomao, Breno Krohling, Igor P Biral, Gabriel G de Angelo, Fábio CR Alves Jr, José GM Esgario, Alana C Simora, Pedro BC Castro, et al. 2020. PAD-UFES-20: A skin lesion dataset composed of patient data and clinical images collected fr...

    2020

  27. [35]

    Sabela Paradela-De-La-Morena, Rosa Fernandez-Torres, Walter Martínez-Gómez, and Eduardo Fonseca-Capdevila. 2015. Teledermatology: diagnostic reliability in 383 children. European Journal of Dermatology 25 (2015), 563–569

    2015

  28. [36]

    Paola Pasquali. 2014. Photography in Dermatology. In Skin Cancer . https: //api.semanticscholar.org/CorpusID:74385942

    2014

  29. [37]

    Nikolay Ponomarenko, Lina Jin, Oleg Ieremeiev, Vladimir Lukin, Karen Egiazar- ian, Jaakko Astola, Benoit Vozel, Kacem Chehdi, Marco Carli, Federica Battisti, et al. 2015. Image database TID2013: Peculiarities, results and perspectives.Signal Enhanced Dermatology Image Quality ...

    2015

  30. [38]

    Nikolay Ponomarenko, Vladimir Lukin, Alexander Zelensky, Karen Egiazarian, Marco Carli, and Federica Battisti. 2009. TID2008-a database for evaluation of full- reference visual quality assessment metrics. Advances of modern radioelectronics 10, 4 (2009), 30–45

    2009

  31. [39]

    G Romero, D De Argila, L Ferrandiz, MP Sánchez, S Vañó, R Taberner, P Pasquali, C de la Torre, F Alfageme, J Malvehy, et al. 2018. Modelos de práctica de la teleder- matología en España. Estudio longitudinal 2009-2014. Actas Dermo-Sifiliográficas 109, 7 (2018), 624–630

    2018

  32. [40]

    Jayesh Ruikar and Saurabh Chaudhury. 2023. NITS-IQA Database: A New Image Quality Assessment Database. Sensors 23, 4 (2023), 2279

    2023

  33. [41]

    Hamid R Sheikh, Muhammad F Sabir, and Alan C Bovik. 2006. A statistical evaluation of recent full reference image quality assessment algorithms. IEEE Transactions on image processing 15, 11 (2006), 3440–3451

    2006

  34. [42]

    Ernestasia Siahaan, Alan Hanjalic, and Judith A Redi. 2016. Does visual quality depend on semantics? A study on the relationship between impairment annoy- ance and image semantics at early attentive stages. Electronic Imaging 28 (2016), 1–9

    2016

  35. [43]

    Leslie N Smith. 2017. Cyclical learning rates for training neural networks. In2017 IEEE winter conference on applications of computer vision (W ACV). IEEE, 464–472

    2017

  36. [44]

    Leslie N Smith and Nicholay Topin. 2019. Super-convergence: Very fast training of neural networks using large learning rates. InArtificial intelligence and machine learning for multi-domain operations applications , Vol. 11006. SPIE, 369–386

    2019

  37. [45]

    Shaolin Su, Hanhe Lin, Vlad Hosu, Oliver Wiedemann, Jinqiu Sun, Yu Zhu, Hantao Liu, Yanning Zhang, and Dietmar Saupe. 2023. Going the extra mile in face image quality assessment: A novel database and model.IEEE Transactions on Multimedia (2023)

    2023

  38. [46]

    Hossein Talebi and Peyman Milanfar. 2018. NIMA: Neural image assessment. IEEE transactions on image processing 27, 8 (2018), 3998–4011

    2018

  39. [47]

    Mingxing Tan. 2019. Efficientnet: Rethinking model scaling for convolutional neural networks. arXiv preprint arXiv:1905.11946 (2019)

    2019 arXiv

  40. [48]

    Nello Tommasino, Matteo Megna, Sara Cacciapuoti, Alessia Villani, Fabrizio Martora, Angelo Ruggiero, Lucia Genco, and Luca Potestio. 2024. The Past, the Present and the Future of Teledermatology: A Narrative Review. Clinical, Cosmetic and Investigational Dermatology Volume 17 ...

    2024 doi

  41. [49]

    S Vañó-Galván, A Hidalgo, I Aguayo-Leiva, M Gil-Mosquera, L Ríos-Buceta, MN Plana, J Zamora, A Martorell-Calatayud, and P Jaén. 2011. Store-and-forward teledermatology: assessment of validity in a series of 2000 observations. Actas Dermo-Sifiliográficas (English Edition) 102, ...

    2011

  42. [50]

    Maria João M Vasconcelos, Luís Rosado, and Márcia Ferreira. 2014. Principal axes-based asymmetry assessment methodology for skin lesion image analysis. In International symposium on visual computing . Springer, 21–31

    2014

  43. [51]

    Toni Virtanen, Mikko Nuutinen, Mikko Vaahteranoksa, Pirkko Oittinen, and Jukka Häkkinen. 2014. CID2013: A database for evaluating no-reference image quality assessment algorithms. IEEE Transactions on Image Processing 24, 1 (2014), 390–402

    2014

  44. [52]

    Kailas Vodrahalli, Justin Ko, Albert S Chiou, Roberto Novoa, Abubakar Abid, Michelle Phung, Kiana Yekrang, Paige Petrone, James Zou, and Roxana Daneshjou

  45. [53]

    Xiaoping Wu, Ni Wen, Jie Liang, Yu-Kun Lai, Dongyu She, Ming-Ming Cheng, and Jufeng Yang. 2019. Joint acne image grading and counting via label distribution learning. In Proceedings of the IEEE/CVF international conference on computer vision. 10642–10651

    2019

  46. [54]

    Jingtao Xu, Peng Ye, Qiaohong Li, Haiqing Du, Yong Liu, and David Doermann

  47. [55]

    Dan Yang, Veli-Tapani Peltoketo, and Joni-Kristian Kamarainen. 2019. CNN- based cross-dataset no-reference image quality assessment. In Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops . 0–0

    2019

  48. [56]

    Weixia Zhang, Kede Ma, Guangtao Zhai, and Xiaokang Yang. 2021. Uncertainty- aware blind image quality assessment in the laboratory and wild. IEEE Transac- tions on Image Processing 30 (2021), 3474–3486

    2021

  49. [57]

    Timothy Zoltie, Sigrid Blome-Eberwein, Sarah Forbes, Mike Theaker, and Walayat Hussain. 2022. Medical photography using mobile devices. bmj 378 (2022). Received 16 January 2025; revised 26 February 2025; accepted 26 February 2025

    2022

  50. [2016]

    IEEE Transactions on Image Processing 25 (9 2016), 4444–4457

    Blind Image Quality Assessment Based on High Order Statistics Aggre- gation. IEEE Transactions on Image Processing 25 (9 2016), 4444–4457. Issue 9. https://doi.org/10.1109/TIP.2016.2585880

    2016

  51. [2017]

    In Design and Quality for Biomedical Technologies X , Vol

    Image quality assessment for teledermatology: from consumer devices to a dedicated medical device. In Design and Quality for Biomedical Technologies X , Vol. 10056. SPIE, 105–112

  52. [2020]

    American journal of clinical dermatology 21 (2020), 41–47

    Artificial intelligence in dermatology—where we are and the way to the future: a review. American journal of clinical dermatology 21 (2020), 41–47

    2020

  53. [2022]

    arXiv preprint arXiv:2209.09105 (2022)

    Development and Clinical Evaluation of an AI Support Tool for Improving Telemedicine Photo Quality. arXiv preprint arXiv:2209.09105 (2022)

    2022 arXiv

  54. [2023]

    Journal of the American Academy of Dermatology 88, 4 (2023), 927–928

    Dermatology Image Quality Assessment (DIQA): Artificial intelligence to ensure the clinical utility of images for remote consultations and clinical trials. Journal of the American Academy of Dermatology 88, 4 (2023), 927–928

    2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.