About

I am a Royal Society-Newton Fellow at University of Hertfordshire, UK (https://biomachinelearning.net/). I am also associated  with CSIR-Central Scientific Instrumentation Organisation, Chandigarh, India (https://www.csio.res.in/) as a Scientist.

Google Scholar Profile

Github profile

Research

My research  interests  and projects revolve around three categories.

• Understanding Chemical space of odorants

Fragrance and aroma are an essential part of the animal kingdom. Yet, the relationship between the chemical structure of odorants and their percepts remain largely elusive. There are many unanswered questions in this domain in terms of clear identification, differentiation and mathematical definition of the olfactory space. I want to understand the concept of olfactory space, the concept of primaries if at all there is one in this space,  and the interaction of physical and receptor space. For this, I intend to use machine learning and neuroscientific techniques. I  collaborate with people in neuroscience and mathematics in this research.

• • Predicting human olfactory perception from chemical features of odor molecules
  • Understanding the Odour Spaces: A Step towards Solving Olfactory Stimulus-Percept Problem
  • DeepFragrance a peek into the olfactory chemical space

• Odor Source Localisation Techniques

Odor source localisation has been studied extensively in biology, e.g. moth mate seeking, lobster foraging, mosquito host tracking. Research in autonomous, robotic odor localisation has mimicked these behaviours. It is relevant for detecting wildfires, oil spills, industrial gas leaks, and for search and rescue in disaster management etc. I work on designing better algorithms/techniques using existing hardware and different sensing modalities.  I collaborate with people from robotics, sensor development in this area.

• • Consensus-Based Odor Source Localization by Multiagent Systems Under Resource Constraints
  • Cooperative control of multi-agent systems to locate source of an odor
  • Consensus-Based Odor Source Localisation by Multiagent Systems

• Electronic Nose and Tongue system development

I have a keen interest in electronic nose and tongue system development which can be used for various application such as quality quantification of food items and early diagnosis of plant pest attack. I have worked with various companies in this area.

• • A Simple Electronic Tongue
  • Circuit arrangement to suppress crosstalk in chemo-resistive sensor arrays
  • Towards biological plausibility of electronic noses: A spiking neural network based approach for tea odour classification

I have also worked on  the design of human opinion dynamics based optimization algorithm. This work was published in Nature, Scientific Reports and the software is open sourced here.

This space will mainly have my recent work and some interesting thoughts that I may put up time to time.

Detailed CV  here