Year-2023 http://repository.iiitd.edu.in/xmlui/handle/123456789/1316 2026-04-11T04:30:52Z 2026-04-11T04:30:52Z Bhadauriya, Harsh Vardhan Jerripothula, Koteswar Rao (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1692 2025-07-08T22:00:23Z 2023-05-01T00:00:00Z

Skeleton-based interactive object co-part segmentation Bhadauriya, Harsh Vardhan; Jerripothula, Koteswar Rao (Advisor) Object co-part segmentation, which involves segmenting shared objects into meaningful parts in a group of images, is a challenging joint-processing task. Although fully unsupervised deep learning algorithms exist for this task, the resultant parts often lack semantic meaning. This is because these algorithms use latent space to separate the parts, which may not necessarily correspond to meaningful parts as perceived by humans. Additionally, the number of parts required by these algorithms is difficult to pre-determine due to pose and size variations shared objects may exhibit across images, making human interaction necessary. While some interactive methods exist, none of them have explored the use of skeletons, which provide an object structure that can be leveraged to generate meaningful parts. Our proposed approach addresses this gap by presenting a skeleton-based interactive co-part segmentation framework that draws benefits from both unsupervised deep learning and human interaction. The framework employs the correspondence capabilities offered by deep learning counterparts and utilizes skeletons to generate meaningful parts. Experiments on Pascal-Part dataset demonstrate that our proposed framework outperforms existing interactive co-part segmentation methods in terms of segmentation accuracy and meaningfulness of parts.

2023-05-01T00:00:00Z Sharma, Anjali Sethi, Tavpritesh (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1689 2026-01-27T22:00:51Z 2023-05-01T00:00:00Z

Advancing gene signature discovery with generative models : a case study in sepsis Sharma, Anjali; Sethi, Tavpritesh (Advisor) Sepsis and diabetes present intricate medical conditions that present substantial challenges to healthcare systems globally. Timely detection and precise diagnosis are critical in facilitating effective treatment and enhancing patient outcomes. This study aimed to investigate the potential of machine learning-based data augmentation approaches for biomarker discovery in a multicenter dataset of patients with sepsis and diabetes. Specifically, the study focused on comparing the efficacy of three different approaches, including the Gaussian Mixture Model (GMM), Bayesian Network (BN), and Conditional Tabular Generative Adversarial Network (CTGAN), for augmenting microarray expression data in the XpressionSuite tool developed by the TavLab at IIIT-Delhi. Differential Gene Expression Analysis (DGEA) was performed on the augmented data, and statistical significance was compared across the three approaches. The findings indicated that CTGAN-generated data exhibited higher statistical significance than the other two approaches, making it the preferred choice for further analysis. Interestingly, Myc targets were identified as a hallmark in all the models, suggesting the potential involvement of Myc in sepsis in patients with diabetes. Furthermore, the DEGs identified through CTGAN-based DGEA were subjected to functional enrichment analysis. The findings highlighted the involvement of several cytosolic components, including secretory vesicles, secretory granules, and dysregulation of stem cell differentiation, in the pathogenesis of sepsis in patients with diabetes. The study results underscore the potential of data augmentation in enhancing the statistical power of gene expression data analysis. Moreover, the study findings suggest that CTGAN-based data augmentation could be a promising approach for biomarker discovery in patients with sepsis and diabetes. The identified immune system pathways could also serve as potential targets for developing therapeutic interventions for sepsis in diabetic patients. It provides insights into the effectiveness of different data augmentation approaches and their potential for biomarker discovery in sepsis patients with diabetes, with potential implications for advancing clinical research in this area.

2023-05-01T00:00:00Z Kottur, Shaguftha Zuveria Shah, Rinku (Advisor) Tammana, Praveen (Advisor) Kadiyala, Krishna (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1358 2023-12-19T22:00:17Z 2023-05-01T00:00:00Z

Implementing ChaCha-based crppto primitives on programmable smartNICs Kottur, Shaguftha Zuveria; Shah, Rinku (Advisor); Tammana, Praveen (Advisor); Kadiyala, Krishna (Advisor) Control and management plane applications such as serverless function orchestration and 4G/5G control plane functions are offloaded to smartNICs to reduce communication and processing latency. Such applications involve multiple inter-host interactions that were traditionally secured using SSL/TLS gRPC-based communication channels. Offloading the applications to smartNIC implies the security algorithms must also be offloaded. Otherwise, there is a need to send the application messages to the host VM/container for crypto operations, negating offload benefits. This work proposes crypto externs for Netronome Agilio smartNICs that implements authentication and confidentiality (encryption/decryption) using the ChaCha stream cipher algorithm. AES and ChaCha are two popular cipher suites, but ChaCha was chosen since none of the smartNICs have ChaCha-based crypto accelerators. However, smartNICs have restricted instruction set, and limited memory, making it difficult to implement security algorithms. This work identifies and addresses several challenges to implement ChaCha crypto primitives successfully. The evaluations show that the proposed crypto extern implementation satisfies the scalability requirement of popular applications such as serverless management functions and host in-band network telemetry.

2023-05-01T00:00:00Z Monteiro, Kyzyl John Parnami, Aman (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1357 2023-12-19T22:00:19Z 2023-05-01T00:00:00Z

Enabling rapid prototyping of tangible augmented reality experiences Monteiro, Kyzyl John; Parnami, Aman (Advisor) We introduce Teachable Reality, an augmented reality (AR) prototyping tool for creating interactive tangible AR applications with arbitrary everyday objects. Teachable Reality leverages vision-based interactive machine teaching (e.g., Teachable Machine), which captures real-world interactions for AR prototyping. It identifies the user-defined tangible and gestural interactions using an on-demand computer vision model. Based on this, the user can easily create functional AR prototypes without programming, enabled by a trigger-action authoring interface. Therefore, our approach allows the flexibility, customizability, and generalizability of tangible AR applications that can address the limitation of current marker-based approaches. We explore the design space and demonstrate various AR prototypes, which include tangible and deformable interfaces, context-aware assistants, and body-driven AR applications. The results of our user study and expert interviews confirm that our approach can lower the barrier to creating functional AR prototypes while also allowing flexible and general-purpose prototyping experiences.

2023-05-01T00:00:00Z