Priyadarshini Kumari
Sony Research Previously @IIT Bombay
I am a Research Scientist at Sony Research.
My research aims to create scalable, data-efficient ML models that continuously learn, adapt, and ensure interpretability in decision processes. My work has explored the applicability of these models in various domains, including biomedical research for scientific discoveries, the design of recommendation systems, and the development of multimodal perception models encompassing vision, speech, text, and olfactory inputs.
Previously, I received my Ph.D. from IIT Bombay, advised by Prof. Subhasis Chaudhuri and Prof. Siddhartha Chaudhuri. My thesis was on Label-Efficient Distance Metric Learning. Before that, I completed my master’s also from IIT Bombay. As a part of my master’s thesis, I developed multimodal rendering techniques to synthesize a combined hapto-visual-auditory perceptual experience of interaction with 3D models of heritage sites. The goal was to provide access to heritage objects to visually-impaired people.
Here is my CV
news
Oct 31, 2023 | Our paper “FRUNI and FTREE synthetic knowledge graphs for evaluating explainability” is accepted at NeurIPS XAIA workshop 2023. |
---|---|
Sep 21, 2023 | Two papers “Perceptual metrics for odorants: learning from non-expert similarity feedback using machine learning” and “Comparing molecular representations, e-nose signals, and other featurization, for learning to smell aroma molecules” are accepted at PLOS One |
Jul 15, 2023 | Our paper “Optimizing Learning Across Multimodal Transfer Features for Modeling Olfactory Perception” was accepted to Multimodal SIGKDD 2023 |
Jul 12, 2023 | I gave a talk on “Using the dynamics of discovery: A temporal graph-based approach to automated hypothesis generationat” at 3rd Nobel Turing Challenge Initiative Workshop |
Apr 28, 2023 | I will serve as senior program chair for WiML un-workshop @ ICML 2023 |
selected publications
-
- Optimizing Learning Across Multimodal Transfer Features for Modeling Olfactory PerceptionIn International Workshop on Multimodal Learning at SIGKDD , 2023
We introduce a novel multilabel and multimodal transfer learning technique for modeling olfactory perception. Our approach aims to tackle the challenges of data scarcity and label skewness in the olfactory domain.
- PerceptNet: Learning perceptual similarity of haptic textures in presence of unorderable tripletsIn IEEE World Haptics Conference (WHC) , 2019
We propose a deep metric learning approach that demonstrates the utility of ambiguous triplets for effectively modeling human perception.