About
PhD Candidate at IIIT-Delhi
I'm a fourth-year Ph.D. candidate at IIIT-Delhi, exploring the fascinating intersection where technology meets human emotions. My research bridges multiple disciplines: artificial intelligence, ubiquitous computing, human-computer interaction, and healthcare.
Imagine a world where your wearable device not only tracks your steps but understands your emotional state. I'm developing predictive models that decode the subtle language of physiological signals—your heartbeat, skin conductance, body temperature—to assess mental well-being in everyday settings. This interdisciplinary approach allows me to build AI systems that are both technically robust and deeply human-centered.
What makes my work unique is its focus on the complexities of emotion data through an HCI lens. By understanding how humans experience, express, and report emotions, I design algorithms that can recognize emotional patterns more accurately in real-world environments—where data is messy, context matters, and individual differences abound.
Research Interests
Professional Activities
Reviewer: CHI, CSCW, IMWUT, UbiComp, Percom, WiML workshop (Neurips)
TA: Interactive Systems, Mobile Computing, Research Methods, Computer Networks
Volunteering
ACM Compass 2024 Organizing Team
TinyML India Organizing Team
Program Co-chair: AutoMLPerSys 2025
Steering Committee: AutoMLPerSys 2024
Publications
EEVR: A Dataset of Paired Physiological Signals and Textual Descriptions for Joint Emotion Representation Learnings
Authors: Pragya Singh, Ritvik Budhiraja, Ankush Gupta, Anshul Goswami, Mohan Kumar, Pushpendra Singh
The EEVR (Emotion Elicitation in Virtual Reality) dataset is a novel resource created for language-supervision-based pre-training and emotion recognition tasks, such as classifying valence and arousal. It includes high-quality physiological signals paired with qualitative textual descriptions of emotions. We evaluated the dataset using the Contrastive Language Signal Pre-training (CLSP) method, which combines physiological signals with self-reported emotional annotations. This approach significantly improved performance in emotion recognition, with a 20% increase in arousal classification and a 10% increase in valence classification.
Translating Emotions to Annotations: A Participant's Perspective of Physiological Emotion Data Collection
Authors: Pragya Singh, Ritvik Budhiraja, Mohan Kumar, Pushpendra Singh
Physiological signals hold immense potential for ubiquitous emotion monitoring, presenting numerous applications in emotion recognition. However, harnessing this potential is hindered by significant challenges, particularly in the collection of annotations that align with physiological changes since the process hinges heavily on human participants. In this work, we set out to study the perspectives of human participants involved in the emotion data collection procedure using 360° virtual reality video stimulus followed by semi-structured interviews.
"But I Won't Say That It Was Bad Seeing a Real Vagina": Understanding Perspectives toward Learning Sensitive-Critical Health Topic
Authors: Sara Moin, Manshul Belani, Pragya Singh, Nishtha Phutela, Pushpendra Singh
In India, topics related to sexual and reproductive health (SRH) are rarely discussed openly due to stigma. To understand the attitudes towards SRH, we designed a Cervical cancer awareness tutorial in Virtual Reality and collected data from 66 participants across genders and life stages through interviews, self-reported emotions, and physiological sensor data. Our findings revealed an acute lack of knowledge about self-body anatomy and a need for creating health literacy.
Can we say a cat is a cat? Understanding the challenges in annotating physiological signal-based emotion data
Authors: Pragya Singh, Mohan Kumar, Pushpendra Singh
This paper presents a position discussion on the current technique of annotating physiological signal-based emotion data. Our discourse underscores the importance of adopting a nuanced understanding of annotation processes, paving the way for a more insightful exploration of the intricate relationship between physiological signals and human emotions.
Generating Tiny Deep Neural Networks for ECG Classification on MicroControllers
Authors: S. Mukhopadhyay, S. Dey, A. Ghose, Pragya Singh and P. Dasgupta
This paper shows that Neural Architecture Search (NAS) can be used to generate tiny but accurate multi-objective models for classifying ECG signals. Our framework is the first of its kind for automatically generating a DNN for screening Atrial Fibrillation on an MCU. Moreover, our research shows that the proposed NAS finds more accurate tiny models than human-designed ones and is effective in enabling customized solutions for a resource-limited target platform.
Awards and Fellowships
Chanakya Doctoral Fellowship
From Ihub Anubhuti (August 2024)
Microsoft Conference Travel Grant
For NeurIPS 2024 (September 2024)
Award Finalist
For Poster Presentation at The Machine Learning Summer School in Okinawa 2024 (March 2024)
Contact
If our research interests align, I would be grateful to connect and collaborate on new ideas.
Location:
IIITD, Delhi, India
Email:
pragyas@iiitd.ac.in
pragyasingh18rathore@gmail.com