Search

Your search keyword '"Singh, Sonam"' showing total 289 results
Start Over Author "Singh, Sonam"
289 results on '"Singh, Sonam"'

3. Automatic Discovery and Assessment of Interpretable Systematic Errors in Semantic Segmentation

Author

Singh, Jaisidh, Singh, Sonam, Kale, Amit Arvind, and Gandhi, Harsh K

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

This paper presents a novel method for discovering systematic errors in segmentation models. For instance, a systematic error in the segmentation model can be a sufficiently large number of misclassifications from the model as a parking meter for a target class of pedestrians. With the rapid deployment of these models in critical applications such as autonomous driving, it is vital to detect and interpret these systematic errors. However, the key challenge is automatically discovering such failures on unlabelled data and forming interpretable semantic sub-groups for intervention. For this, we leverage multimodal foundation models to retrieve errors and use conceptual linkage along with erroneous nature to study the systematic nature of these errors. We demonstrate that such errors are present in SOTA segmentation models (UperNet ConvNeXt and UperNet Swin) trained on the Berkeley Deep Drive and benchmark the approach qualitatively and quantitatively, showing its effectiveness by discovering coherent systematic errors for these models. Our work opens up the avenue to model analysis and intervention that have so far been underexplored in semantic segmentation., Comment: 7 pages main paper (without references), total 13 pages & 9 figures

Published
2024

5. Concrete Strength Prediction Using Machine Learning

Author

Vairagar, Shubhangi, Khilare, Samarth Pravin, Gunjal, Azad Navnath, Bhaisare, Mohit Vivek, Kalingan, Harsh Mahesh, Singh, Sonam, Shinde, Sarika, Li, Gang, Series Editor, Filipe, Joaquim, Series Editor, Xu, Zhiwei, Series Editor, Singh, Rajesh, editor, and Gehlot, Anita, editor

Published
2025
Full Text
View/download PDF

9. Linguistic Elements of Engaging Customer Service Discourse on Social Media

Author

Singh, Sonam and Rios, Anthony

Subjects
Computer Science - Computation and Language
Abstract

Customers are rapidly turning to social media for customer support. While brand agents on these platforms are motivated and well-intentioned to help and engage with customers, their efforts are often ignored if their initial response to the customer does not match a specific tone, style, or topic the customer is aiming to receive. The length of a conversation can reflect the effort and quality of the initial response made by a brand toward collaborating and helping consumers, even when the overall sentiment of the conversation might not be very positive. Thus, through this study, we aim to bridge this critical gap in the existing literature by analyzing language's content and stylistic aspects such as expressed empathy, psycho-linguistic features, dialogue tags, and metrics for quantifying personalization of the utterances that can influence the engagement of an interaction. This paper demonstrates that we can predict engagement using initial customer and brand posts., Comment: Accepted to NLP+CSS at EMNLP 2022

Published
2022

13. SAFETY SHOULD BE A PERFORMANCE DRIVER.

Author

Mittal, Vikas, Piazza, Alessandro, and Singh, Sonam

Subjects
BUSINESS enterprises, SAFETY, PRODUCT safety, INDUSTRIAL safety, ORGANIZATIONAL performance
Abstract

Safety is regarded as an indispensable right for customers and employees. Government agencies exist to enforce standards, and firms spend millions testing their products and creating safe workplace environments. And yet products are frequently recalled, and workplace accidents continue to happen. Why aren't companies doing better on safety? Most executives frame safety as a compliance issue. They see it as a cost and, consequently, they underinvest in it. They tend to treat safety as an abstract value rather than as a driver of performance. And when a safety crisis hits, they often react with unsustainable measures, generally aimed at managing their public image. To help companies get out of this rut, the authors present evidence that safety can be a key driver of performance. Then they offer a five-step process for leaders: align on the definition of safety, agree on which metrics to use, anticipate and prevent problems, customize safety training, and incentivize employees to adopt preventive behaviors. By reimagining safety not as a defensive necessity but as an offensive opportunity, companies can elevate safety from a siloed function to a shared mindset, and from a cost center to a value accelerator. [ABSTRACT FROM AUTHOR]

Published
2024

14. The Value of Safety Training for Business-to-Business Firms.

Author

Chen, Yixing, Sridhar, Shrihari, Han, Kyuhong, Singh, Sonam, Mittal, Vikas, and Im, Taehoon

Subjects
BUSINESS enterprises, WORK-related injuries, SAFETY education, INDUSTRIAL marketing, HAZARD mitigation, CONSTRUCTION workers
Abstract

Business-to-business suppliers invest in safety training programs believing that such programs mitigate safety hazards, prevent workplace injuries, and create value for their customers. However, causal evidence of these effects is sparse. Study 1 uses site-level monthly data from a global oil field services company. Exploiting sharp discontinuities in safety training hours due to catastrophic accidents, the authors find that a 10% increase in safety training hours per capita decreases safety hazards per capita by 6.45%–9.57%. Study 2 measures the causal impact of business establishments' safety training intensity on their workplace injuries: it leverages Local Law 196 requiring workers at construction establishments in New York City to complete at least ten hours of safety training. This legislation reduced injury rates at construction establishments in New York City by.54–.68 percentage points (a 15.56%–18.84% decrease) relative to their counterparts. Study 3, a stated-choice conjoint experiment of business-to-business procurement professionals, documents that the focal supplier's investment in safety training increases the probability of its proposal being selected by those professionals. Collectively, these findings validate the need for suppliers to invest in safety training as a risk-mitigation vehicle that has positive implications for business-to-business buying decisions. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

16. Performance Analysis of 24-Tupled RoF System Based on Parallel MZMs and SOA

Author

Singh, Sonam, Rani, Ankita, Kedia, Deepak, Kacprzyk, Janusz, Series Editor, Gomide, Fernando, Advisory Editor, Kaynak, Okyay, Advisory Editor, Liu, Derong, Advisory Editor, Pedrycz, Witold, Advisory Editor, Polycarpou, Marios M., Advisory Editor, Rudas, Imre J., Advisory Editor, Wang, Jun, Advisory Editor, Singh, Sri Niwas, editor, Mahanta, Saurov, editor, and Singh, Yumnam Jayanta, editor

Published
2023
Full Text
View/download PDF

17. Breast Cancer Prediction Using Auto-Encoders

Author

Singh, Sonam Jawahar, Rajaraman, Ramanathan, Verlekar, Tanmay Tulsidas, Xhafa, Fatos, Series Editor, Goswami, Saptarsi, editor, Barara, Inderjit Singh, editor, Goje, Amol, editor, Mohan, C., editor, and Bruckstein, Alfred M., editor

Published
2023
Full Text
View/download PDF

21. Machine learning based genome cancer dataset approaches: Review.

Author

Singh, Sonam Jawahar and Agrawal, Pooja

Subjects
*DIGITAL technology, *DISEASE risk factors, *MACHINE learning, *DISRUPTIVE innovations, *NUCLEOTIDE sequencing
Abstract

The elementary healthcare system for cancer patients focused on a single medicine that suited many patients. In recent times, the biggest challenge in healthcare is that medicines should be based on disruptive technologies instead of generalized ones. This paper provides a review of various genome cancer dataset approaches using machine learning. A huge amount of data has been generated by patient health monitors, new biotechnology techniques, affordable genome sequencing, and aggregating information on patients' journeys. Precise medication based on the genetic characteristics of the patient and the cancer is needed to administer the perfect therapy at the ideal time to the ideal patient with the ideal malignancy. A doctor can no longer assess all of those data or stay current due to the digital revolution in health. Data scientists must emphasize the significance of developing efficient algorithms based on physicians' knowledge of biomedical imaging and extensive data. Genomics studies genomes' structure, function, evolution, and mapping, which aims to represent and quantify genes. The literature survey suggests that the Machine Learning approach to integrating the datasets provides benefits in cancer research like predictions of diagnosis, prognosis, survival time, patient risk factors, etc. The paper provides a performance and accuracy comparison of the various machine learning techniques for cancer datasets. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

33. Transethnic analysis of psoriasis susceptibility in South Asians and Europeans enhances fine mapping in the MHC and genome wide

Author

Stuart, Philip E., Tsoi, Lam C., Nair, Rajan P., Ghosh, Manju, Kabra, Madhulika, Shaiq, Pakeeza A., Raja, Ghazala K., Qamar, Raheel, Thelma, B.K., Patrick, Matthew T., Parihar, Anita, Singh, Sonam, Khandpur, Sujay, Kumar, Uma, Wittig, Michael, Degenhardt, Frauke, Tejasvi, Trilokraj, Voorhees, John J., Weidinger, Stephan, Franke, Andre, Abecasis, Goncalo R., Sharma, Vinod K., and Elder, James T.

Published
2022
Full Text
View/download PDF

37. A Fully Convolutional Neural Network based Structured Prediction Approach Towards the Retinal Vessel Segmentation

Author

Dasgupta, Avijit and Singh, Sonam

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Automatic segmentation of retinal blood vessels from fundus images plays an important role in the computer aided diagnosis of retinal diseases. The task of blood vessel segmentation is challenging due to the extreme variations in morphology of the vessels against noisy background. In this paper, we formulate the segmentation task as a multi-label inference task and utilize the implicit advantages of the combination of convolutional neural networks and structured prediction. Our proposed convolutional neural network based model achieves strong performance and significantly outperforms the state-of-the-art for automatic retinal blood vessel segmentation on DRIVE dataset with 95.33% accuracy and 0.974 AUC score.

Published
2016

40. Modulation of Microbiome Through Seed Bio-priming

Author

Sarkar, Deepranjan, Chattopadhyay, Arghya, Singh, Sonam, Devika, O. Shiva, Pal, Subhadip, Parihar, Manoj, Pal, Sumita, Singh, Harikesh Bahadur, Rakshit, Amitava, Varma, Ajit, Series Editor, Manoharachary, Chakravarthula, editor, and Singh, Harikesh Bahadur, editor

Published
2020
Full Text
View/download PDF

41. An Efficient Network Coded Routing Protocol for Delay Tolerant Network

Author

Sakle, Mukesh, Singh, Sonam, Kacprzyk, Janusz, Series Editor, Gomide, Fernando, Advisory Editor, Kaynak, Okyay, Advisory Editor, Liu, Derong, Advisory Editor, Pedrycz, Witold, Advisory Editor, Polycarpou, Marios M., Advisory Editor, Rudas, Imre J., Advisory Editor, Wang, Jun, Advisory Editor, Shukla, Rajesh Kumar, editor, Agrawal, Jitendra, editor, Sharma, Sanjeev, editor, Chaudhari, Narendra S., editor, and Shukla, K. K., editor

Published
2020
Full Text
View/download PDF

42. Anti-plane waves in a liquid-loaded piezo-flexo-electric layered model with interface energy.

Author

Singh, Sonam and Singh, Abhishek K

Subjects
*DISPERSION relations, *VISCOELASTIC materials, *PHASE velocity, *WAVENUMBER, *FREE surfaces
Abstract

The basic design of a Love wave (LW) bio-sensor contains the loading of a viscoelastic liquid on top of a layered structure with distinct viscoelastic properties. Changes in the characteristics of the propagating acoustic wave caused by biochemical interactions at the sensing area can be detected at the output Inter digital transducer (IDT). The propagation of the anti-plane (AP) wave is discussed in this work in a layered structure of piezo-flexo-electric (PFE) layer bonded with a PFE half-space having a soft reinforced layer at the interface. The free surface of the PFE layer is loaded with viscous liquid. The viscosity of the loaded liquid introduces losses and results in the damping of the wave. For the formulation of the problem with interface energy, the Gurtin–Murdoch approach is used. Using suitable conditions, a dispersion relation for propagating waves is derived in complex form. On separating the dispersion relation in real and non-real parts, the expressions relating the phase and damp velocities with wave number are derived. The obtained theoretical results are portrayed for the numerical data of PFE materials and distinct data of the interface layer. The obtained results are validated with pre-existing literature. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

45. IN SICHERHEIT.

Author

Mittal, Vikas, Piazza, Alessandro, and Singh, Sonam

Published
2025

Catalog

Books, media, physical & digital resources
See catalog results