Publications: (Google scholar)
Journal Articles:
1. M. Narwaria, “Deep learning–based image super-resolution in microscopy: Why more pixels do not imply higher resolving ability?”, in Journal of Microscopy, 1–18, 2026.
(While Deep Learning based image super resolution is extremely popular, its practical implications from a systems perspective are incorrectly interpreted and, in most cases, overstated. This is one of the very few papers in literature that provides rigorous and well-grounded arguments on why Deep Learning based image super resolution does not and cannot construct meaningful visual details in images. The arguments in the paper rely on solid signal processing and theory of imaging, and debunk unscientific claims about capabilities of Deep Learning in microscopy)
2. P. Saxena, A. K. Tiwari, M. Narwaria, “Adaptive self-Attention enhanced conditional GAN for image dehazing”, in Pattern Recognition, Volume 177, 2026.
(This paper proposes a new method for haze reduction in images. The proposed method introduces a new attention mechanism such that the network learns how much attention to give to different hazy regions automatically. This contrasts with existing attention methods which usually use fixed attention behavior)
3. K. Datta, A. Bhardwaj and M. Narwaria, “Does Attribute Space-Driven Tactile Features Correspond to Overall Perceptual Similarity of Textures?,” in IEEE Sensors Journal, vol. 26, no. 10, pp. 14586-14594, 2026.
(This paper examines critically if machine learning models trained on computational metrics can predict reliably the perceptual texture similarity. Unlike most existing works, this paper takes a more rigorous and principled approach by using the concept of perturbed feature space. The findings of the paper are somewhat surprising and new since we establish that well trained machine learning model with excellent prediction on test data may not learn the underlying mapping. This highlights fundamental drawbacks with traditional and popular machine learning methods.)
4. K. Dutta, A. Bhardwaj and M. Narwaria. “Haptic-Sound Analysis of Materials’ Clustering Based on Tool Tip Tapping Exploration”, in Multisensory Research, 39.3-5 (2026): 417-432.
(This paper approaches the problem of texture assessment via an unconventional route of unsupervised learning. It proposes and develops perceptual filterbanks using fundamental ideas from signal processing for audio and haptic modalities. These filterbanks are unique and novel since they are designed from perceptual sensitivity of human auditory and haptic senses. Therefore, this work strongly emphasizes the scientific importance of incorporating human perception knowledge in algorithm design)
5. M. Narwaria, “System Design: Then and Now [Humour],” in IEEE Signal Processing Magazine, Sept. 2023.
(If data can supposedly reveal everything, is it time to abandon textbooks and domain-knowledge? One of very few and unique publications appearing in IEEE Signal Processing Magazine Humor Special Section; an unusual and light-hearted take on System Design philosophy in an era dominated by machine learning)
6. M. Narwaria, “Explainable Machine Learning: The importance of a system-centric perspective” in IEEE Signal Processing Magazine, vol. 40, no. 2, pp. 165-172, 2023.
(One of the few papers in literature which explicitly analyzes why a blind data fitting approach, which is unfortunately a trend in today’s era of machine learning, is fundamentally flawed. It also underlines how a system-centric approach can provide meaningful and quantifiable technical insights into a machine learning based black box modeling. Written as a Lecture Note to make it accessible to a set of readers from very diverse backgrounds)
7. M. Narwaria, “Does explainable machine learning uncover the black box in vision applications?”, Image and Vision Computing, Volume 118, 2022, 104353, ISSN 0262-8856.
(With several thousands of papers already published on the topic, is the problem of explainable machine learning for vision applications solved? The answer is an emphatic and an unfortunate no. To that end, this paper provides a critical analysis on why current efforts in explainable machine learning for vision applications are fundamentally limited in terms of tackling the black box issue. This paper was also showcased via IIT Jodhpur Media Outreach in both Hindi and English media, 2022)
8. M. Narwaria, "The Transition From White Box to Black Box: Challenges and Opportunities in Signal Processing Education," in IEEE Signal Processing Magazine, vol. 38, no. 3, pp. 163-173, May 2021, doi: 10.1109/MSP.2021.3050996. (A comprehensive analysis on how possible signal processing education should change in an era of machine learning and equally importantly why machine learning education cannot afford to de-emphasize fundamental signal processing ideas; Special Issue on “Innovation Starts With Education”; Prof. Alan V. Oppenheim as one of Lead Guest Editors)
9. Mythili. K. and M. Narwaria, “Assessment of machine learning-based audiovisual quality predictors: Why uncertainty matters,”, ACM Trans. Multimedia Comput. Commun. Appl., vol. 17, no. 2, Apr. 2021.
(Technical insights into why uncertainty and interaction between the auditory and vision modalities for the task of joint audio-visual quality assessment in the context of immersive multimedia communication systems)
10. M. Narwaria and A. Tatu, "Interval-Based Least Squares for Uncertainty-Aware Learning in Human-Centric Multimedia Systems," in IEEE Transactions on Neural Networks and Learning Systems, vol. 32, no. 11, pp. 5241-5246, Nov. 2021, doi: 10.1109/TNNLS.2020.3025834.
(If the target itself is uncertain, shouldn’t machine learning based optimization benefit from uncertainty? Answer is a yes and this paper presents a novel analytical least squares based framework to explicitly account for uncertainty in data during optimization of machine learning models; one of the few works in machine learning which deals with explicit incorporation of domain knowledge for model optimization)
11. M. Narwaria, "Toward Better Statistical Validation of Machine Learning-Based Multimedia Quality Estimators," in IEEE Transactions on Broadcasting, vol. 64, no. 2, pp. 446-460, June 2018, doi: 10.1109/TBC.2018.2832441. (Why a winner-takes-all strategy for testing and validation of machine learning based multimedia quality predictors is a bad idea? This paper provides strong arguments in favour of a more holistic and grounded validation framework)
12. M. Narwaria, L. Krasula and P. Le Callet, "Data Analysis in Multimedia Quality Assessment: Revisiting the Statistical Tests," in IEEE Transactions on Multimedia, vol. 20, no. 8, pp. 2063-2072, Aug. 2018, doi: 10.1109/TMM.2018.2794266.
(The oft-utilized assumptions underlying statistical tests such as normality etc. are incorrectly interpreted in practical use-cases. This paper revisits the statistical tests from first principles and reveals rigorous insights into what the assumptions mean in the context of practical usage in Multimedia Quality Assessment)
13. R. L. Das and M. Narwaria, “Lorentzian based adaptive filters for impulsive noise environments”, IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 64, no. 6, pp. 1529-1539, June 2017.
(System identification in practically ubiquitous impulse noise is challenging. This paper proposes new algorithms for that purpose and carries out theoretical convergence analysis issues.)
14. L. Krasula, M. Narwaria, K. Fliegel, and P. L. Callet, “Preference of experience in image tone-mapping: Dataset and framework for objective measures comparison”, IEEE Journal of Selected Topics in Signal Processing, vol. 11, no. 1, pp. 64-74, Feb 2017.
(End users find it easier to indicate their preference instead of rating Multimedia content. This paper formalizes the said idea for HDR content and develops framework for comparing objective algorithms.)
15. M. Narwaria, M. P. D. Silva, and P. L. Callet, “HDR-VQM: An objective quality measure for high dynamic range video”, Signal Processing: Image Communication, vol. 35, pp. 46-60, 2015.
(One of the first algorithms specifically designed for HDR video quality measurement by explicit consideration of spatio-temporal factors and display characteristics. Widely cited and also used in MPEG HDR Video Coding activities; cross platform implementation developed in partnership with Apple and Samsung; patent granted for this work U.S. Patent-9794554B1, 2017)
16. D. Mocanu, J. Pokhrel, J. Garella, J. Seppanen, E. Liotou, and M. Narwaria, “No-reference video quality measurement: added value of machine learning”, Journal of Electronic Imaging, vol. 24, no. 6, p. 061208, 2015.
(Is machine learning just a tool for maximizing prediction accuracies all the time? Unlike most other works, this paper demonstrates why and how machine learning can be exploited for predicting variability in ratings, than merely maximizing prediction accuracies.)
17. M. Narwaria, R. Mantiuk, M. P. D. Silva, and P. L. Callet, “HDR-VDP-2.2: A calibrated method for objective quality prediction of high-dynamic range and standard images”, Journal of Electronic Imaging, vol. 24, no. 1, p. 010501, 2015.
(A perceptual HDR image quality prediction algorithm based on modeling of relevant features of the human visual system. A popular, well-cited and reasonably accurate method widely used by industry, researchers and standardization bodies in JPEG-HDR codec testing and validation.)
Conference Articles:
Basant Pandagre and Manish Narwaria, “Upsampling of Personalized HRTF for Spatial Audio Rendering: Why Deep Learning is Problematic?”, To Appear in Proc. of CODS-COMAD 2023, IIT Bombay, 4-7 Jan. 2023. (Competitive Student Travel Grant also awarded to this paper)
Basant Pandagre and Manish Narwaria, “Personalized Spatial Audio Rendering for Augmented and Virtual Reality Applications”, Paper Presentation under the theme of “Experiential interface” at the INAE SERB Youth Conclave held at IIT Jodhpur, 15th -18th September 2022. (No Proceedings Published for this event)
R. Nalla, R. Pinge, M. Narwaria and B. Chaudhury, “Priority Based Functional Group Identification of Organic Molecules using Machine Learning," to appear, Proceedings of the ACM India Joint International Conference on Data Science and Management of Data CoDS-COMAD 18, Goa, India, 2018. (Student authors awarded a Travel Grant)
R. Gupta and M. Narwaria, “DA-IICT at MediaEval 2017: Objective Prediction of Media Interestingness", Proceedings of MediaEval 2017, Dublin, Ireland, 2017.
M. Narwaria, M. P. D. Silva, and P. L. Callet, “Study of high dynamic range video quality assessment," Proc. SPIE, Applications of Digital Image Processing XXXVIII, vol. 9599, p. 95990V, 2015.
L. Krasula, M. Narwaria, K. Fliegel, and P. L. Callet, “Rendering of HDR content on LDR displays: An objective approach," Proc. SPIE, Applications of Digital Image Processing XXXVIII, vol. 9599, p. 95990X, 2015.
L. Krasula, M. Narwaria, K. Fliegel, and P. L. Callet, “Influence of HDR reference on observers preference in tone-mapped images evaluation," 2015 Seventh International Workshop on Quality of Multimedia Experience (QoMEX), pp. 1-6, May 2015.
M. Narwaria, C. Mantel, M. P. D. Silva, P. L. Callet, and S. Forchhammer, “An objective method for high dynamic range source content selection," Sixth International Workshop on Quality of Multimedia Experience (QoMEX), pp. 13-18, Sept 2014.
M. Narwaria, M. P. D. Silva, P. L. Callet, and R. Pepion, “Single exposure vs tone mapped high dynamic range images: A study based on quality of experience," 22nd European Signal Processing Conference (EUSIPCO), pp. 2140-2144, Sept 2014.
A. Hines, P. Kendrick, A. Barri, M. Narwaria, and J. A. Redi, “Robustness and prediction accuracy of machine learning for objective visual quality assessment," 2014 22nd European Signal Processing Conference (EUSIPCO), pp. 2130-2134, Sept 2014.
L. Krasula, M. Narwaria, and P. L. Callet, “An automated approach for tone mapping operator parameter adjustment in security applications," Proc. SPIE,Optics, Photonics, and Digital Technologies for Multimedia Applications III, vol. 9138, p. 913803, 2014.
J.Wang, Y. Fang, M. Narwaria, W. Lin, and P. L. Callet, “Stereoscopic image retargeting based on 3d saliency detection," IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 669-673, May 2014.
M. Narwaria, M. P. D. Silva, P. L. Callet, and R. Pepion, “On improving the pooling in HDR-VDP-2 towards better HDR perceptual quality assessment," Proc. SPIE, Human Vision and Electronic Imaging XIX, vol. 9014, p. 90140N, 2014.
M. Narwaria, M. P. D. Silva, P. Callet and R. Pepion, “Impact of Tone Mapping In High Dynamic Range Image Compression, in Proc. Eighth International Workshop on Video Processing and Quality Metrics for Consumer Electronics (VPQM), 2014.
M. Narwaria, M. P. D. Silva, P. L. Callet, and R. Pepion, “Adaptive contrast adjustment for postprocessing of tone mapped high dynamic range images," IEEE International Symposium on Circuits and Systems (ISCAS2013), pp. 1103-1106, May 2013.
M. Narwaria, M. P. D. Silva, P. L. Callet, and R. Pepion, “Effect of tone mapping operators on visual attention deployment," Proc. SPIE, Applications of Digital Image Processing XXXV, vol. 8499, p. 84990G, 2012.
M. Narwaria and W. Lin, “Machine learning based modeling of spatial and temporal factors for video quality assessment," 2011 18th IEEE International Conference on Image Processing, pp. 2513-2516, Sept 2011. (Best student paper award nomination)
M. Narwaria and W. Lin, “Video quality assessment using temporal quality variations and machine learning," 2011 IEEE International Conference on Multimedia and Expo, pp. 1-6, July 2011. (Designated as Top 15% paper out of 744 submissions)
W. Lin and M. Narwaria, “Perceptual image quality assessment: recent progress and trends," in Proc. SPIE, Visual Communications and Image Processing, 774403, 2010.
M. Narwaria and W. Lin, “Objective Image Quality Assessment with Singular Value Decomposition, in Proc. Eighth International Workshop on Video Processing and Quality Metrics for Consumer Electronics (VPQM), 2010.
M. Narwaria, W. Lin, I. McLoughlin, S. Emmanuel, and C. Tien, “Non-intrusive speech quality assessment with support vector regression," in Advances in Multimedia Modeling (S. Boll, Q. Tian, L. Zhang, Z. Zhang, and Y.-P. Chen, eds.), vol. 5916 of Lecture Notes in Computer Science, pp. 325-335, Springer Berlin Heidelberg, 2010.
M. Narwaria and W. Lin, “Scalable image quality assessment based on structural vectors," 2009 IEEE International Workshop on Multimedia Signal Processing, pp. 1-6, Oct 2009.