Experience

UVA Biocomplexity Institute

• United States
• Research Services
• 1 - 100 Employee

• Research Scientist

  • Jul 2021 - Present

• Research Associate

  • May 2019 - Jul 2021



NCSU ECE Dept

• Raleigh, North Carolina

• Post Doc

  • May 2018 - May 2019

  Worked on change detection problem in image analysis using optimal transport (OT), a versatile tool for comparing data distribution with applications in economics, urban planning, transfer learning strategies (inferring labels for unlabelled data). Developed scalable algorithms with linear complexity for the computationally expensive OT and employed it for a multiscale image analysis strategy. Developed a principled deep neural network (DNN) architecture in collaboration with a… Show more Worked on change detection problem in image analysis using optimal transport (OT), a versatile tool for comparing data distribution with applications in economics, urban planning, transfer learning strategies (inferring labels for unlabelled data). Developed scalable algorithms with linear complexity for the computationally expensive OT and employed it for a multiscale image analysis strategy. Developed a principled deep neural network (DNN) architecture in collaboration with a graduate student for 8x super-resolution in remote sensing images. This work resulted in a conference publication. Show less Worked on change detection problem in image analysis using optimal transport (OT), a versatile tool for comparing data distribution with applications in economics, urban planning, transfer learning strategies (inferring labels for unlabelled data). Developed scalable algorithms with linear complexity for the computationally expensive OT and employed it for a multiscale image analysis strategy. Developed a principled deep neural network (DNN) architecture in collaboration with a… Show more Worked on change detection problem in image analysis using optimal transport (OT), a versatile tool for comparing data distribution with applications in economics, urban planning, transfer learning strategies (inferring labels for unlabelled data). Developed scalable algorithms with linear complexity for the computationally expensive OT and employed it for a multiscale image analysis strategy. Developed a principled deep neural network (DNN) architecture in collaboration with a graduate student for 8x super-resolution in remote sensing images. This work resulted in a conference publication. Show less



Indian Institute of Science (IISc)

• India
• Research
• 700 & Above Employee

• Research Associate

  • May 2017 - May 2018

  Formulated a Bayesian deep deconvolutional neural network (BD2N2) for developing a sparsity-enforcing statistical model and deployed it for localization of fluorophores in stochastic localization microscopy. Obtained a 30\% improvement in signal reconstruction quality (PSNR) compared to existing techniques. Work was in collaboration with a masters student and super-resolution microscopy researchers and resulted in a publication at the NIPS 2017 workshop and the prestigious focus on… Show more Formulated a Bayesian deep deconvolutional neural network (BD2N2) for developing a sparsity-enforcing statistical model and deployed it for localization of fluorophores in stochastic localization microscopy. Obtained a 30\% improvement in signal reconstruction quality (PSNR) compared to existing techniques. Work was in collaboration with a masters student and super-resolution microscopy researchers and resulted in a publication at the NIPS 2017 workshop and the prestigious focus on microscopy conference. Show less Formulated a Bayesian deep deconvolutional neural network (BD2N2) for developing a sparsity-enforcing statistical model and deployed it for localization of fluorophores in stochastic localization microscopy. Obtained a 30\% improvement in signal reconstruction quality (PSNR) compared to existing techniques. Work was in collaboration with a masters student and super-resolution microscopy researchers and resulted in a publication at the NIPS 2017 workshop and the prestigious focus on… Show more Formulated a Bayesian deep deconvolutional neural network (BD2N2) for developing a sparsity-enforcing statistical model and deployed it for localization of fluorophores in stochastic localization microscopy. Obtained a 30\% improvement in signal reconstruction quality (PSNR) compared to existing techniques. Work was in collaboration with a masters student and super-resolution microscopy researchers and resulted in a publication at the NIPS 2017 workshop and the prestigious focus on microscopy conference. Show less



Idiap Research Institute

• Switzerland
• Research
• 100 - 200 Employee

• Researcher

  • Jan 2012 - Jul 2012

  Was awarded the Indo-Swiss Joint Research Fellowship and as a part of which we developed robust features called Gammatone wavelet cepstral coefficients for automatic speech recognition systems and achieved a consistent improvement of 2% in word accuracy over various real-world noise types. Was awarded the Indo-Swiss Joint Research Fellowship and as a part of which we developed robust features called Gammatone wavelet cepstral coefficients for automatic speech recognition systems and achieved a consistent improvement of 2% in word accuracy over various real-world noise types.