http://repository.iiitd.edu.in/xmlui/handle/123456789/1080 2026-04-11T03:41:45Z 2026-04-11T03:41:45Z Arora, Ansh Subramanyam, A V (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1646 2024-07-20T22:00:19Z 2022-05-01T00:00:00Z

Fine-grained unsupervised tracklet matching Arora, Ansh; Subramanyam, A V (Advisor) Re-identification offers a useful tool for non-invasive biometric validation, surveillance, and human-robot interaction in a broad range of applications from crowd traffic management to personalised healthcare. Given the leaps and bounds that Artificial Intelligence has moved forward with applications almost everywhere, the tasks of Person Re-Identification and Vehicle Re-Identification still remain relatively unapplied. The performance of Deep Learning (DL) in the domain of Videos, due to it probably being the most challenging of fields to work with in the field of Computer Vision, still remains relatively low. However with the introduction of new DL based techniques in the field of Unsupervised Representation Learning, along with more video data than ever being created on a daily basis, work on the field is more in demand than ever before. In this project we aim to create a Dataset Pipeline that creates efficient task based datasets in an unsupervised manner, and create a mini dataset for the task of Person Re-Identification (Person Re-ID). We also aim at learning feature representations from these datasets using Unsupervised Representation Learning techniques such as Contrastive Learning, and then Transfer the feature learnings to the task of Person Re-ID.

2022-05-01T00:00:00Z Singh, Arihant Chhabra, Smiti Mohania, Mukesh (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1209 2023-04-16T22:00:25Z 2022-12-01T00:00:00Z

Design of FMS application Singh, Arihant; Chhabra, Smiti; Mohania, Mukesh (Advisor) Facility Management Services, commonly known as FMS, is the body that manages various services provided by the Institute for the welfare of the IIIT-D family. It is mainly responsible for providing a smooth experience to everyone on campus by helping them with their day-today needs. Our project aims to revamp this for both the user and the admin side to increase accessibility and provide a more user-friendly and efficient method.

2022-12-01T00:00:00Z Wadhwa, Pritish Satija, Gitansh Raj Shah, Rajiv Ratn (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1208 2023-04-16T22:00:21Z 2022-07-01T00:00:00Z

Legal text processing Wadhwa, Pritish; Satija, Gitansh Raj; Shah, Rajiv Ratn (Advisor) Legal Domain has been explored by the scientific community only upto a limited extent. While there is some amount of research available in the United States of America, the amount of research in India in the same domain is comparatively very small. Through this research we aim to explore this exciting domain. In this research, we introduce two novel datasets. The first one consists of about 2 million Indian Legal documents spanning over the 74 years since our country’s independence in 1947. The other dataset comprises of about 7 million US case laws, spanning over a little shy of 4 centuries. Along with this, we also aim to tackle two tasks. First one concerns itself with predicting judgement for Indian Supreme Court cases. The other one focuses on the legal citation worthiness of a statement. We hope that through our efforts, the interest of the research community increases in this domain.

2022-07-01T00:00:00Z Kaur, Gursimran Arora, Ishaan Ray, Sonia Baloni (Advisor) Chakrabarty, Mrinmoy (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/1207 2023-04-16T22:00:25Z 2022-05-01T00:00:00Z

Uncanny valley effect : an EEG study Kaur, Gursimran; Arora, Ishaan; Ray, Sonia Baloni (Advisor); Chakrabarty, Mrinmoy (Advisor) Facial expressions are widely used for interpersonal communication as they reflect the internal affective or emotional state of an individual. Humans are highly dependent on facial cues and expressions to judge the effective state of people around them. However, humans react differently when they see a human, a humanoid or a mechanical bot. Mori, 1970 showed that humans respond to androids as a function of their feature similarity to humans and is popularly known as the ‘Uncanny Valley effect’(UVE). UVE pertains to an eerie feeling when we see or interact with robots. Given that facial cues play a significant role in human communication we aim to test if UVE is associated with face processing. An electroencephalogram (EEG) provides us access to the electrical activity in our brain via small, metal discs (electrodes) attached to our scalp. The Event-Related Potential (ERP) of N-170 (Negative potential of 170 msec after stimulus onset) from EEG is a robust marker associated with face processing. The amplitude of the N170 component is found to be greater for human faces as compared to non-face stimuli. So we hypothesize that if the uncanny valley effect is related to face processing, then there would be differential characteristics of the N170 component elicited for the human face as compared to the robotic face processing. To test this hypothesis we asked participants to rate human and robotic faces images shown on computer monitor while recording EEG signals simultaneously. The rating was done on an analog scale for each individual image and was done on the basis of two metrics: Likeability and Mechano-humanness. The likeabiltiy defines how friendly is the face to look at and mechano-humanness indicating how mechanical does the face look like. The range of scale for likeability was from -100 to 100 and the range for mechano-humanness being 0 to 100. Rating data when analyzed showed a trend similar to the uncanny valley effect. Epochs of EEG aligned to the onset of face (human/robotic) stimulus were analyzed offline to study the N170 ERP. The results showed the presence of N170 component for both human and robotic faces. However, the magnitude of the N170 component was larger for human faces as compared to robotic faces. While data collection is underway to substantiate the hypothesis statistically, the preliminary results till date suggest the contribution of face processing regions of the brain towards the UVE.

2022-05-01T00:00:00Z