Industry Projects
- Optimised algorithm design of computer vision applications (object detection/recognition, person recognition) and porting the standard library primitives to TSMC 5nm process library.
- Designing an efficient pipeline for distant person detection following recognition in 4K video using GPUs.
- A novel deep neural design and efficient Pipeline architecture for Person Re-Identification in high resolution Video.
- Fragmented Huffman-Based Compression Methodology for CNN Targeting Resource-Constrained Edge Devices. (A novel compression and decompression methodology applied on binary equivalent of a numeric file).
- Developed a novel approach in WIFI-CSI based person fall detection.
- Explored the MLPerf neural network models in optimising the various operations of the network algorithms for porting into custom neural engine (applications ranging from vision classification, object detection, natural language processing).
- Design a custom transformer-based graph neural network (GNN) for ecommerce based recommendation system.
- Additionally taught AI/ML courses as a part of faculty development programs across few universities as industry resource person.
Semiconductor Industry Experience
AI Research Engineer
August 2019 - August 2021
Redpine Signals Inc (Acquired by Silicon Labs)
Senior Engineer
September 2021 - April 2022
Ceremorphic Technologies India Pvt Ltd (erstwhile Redpine Signals)
Worked on neural net accelerator design.
DRDO Project Delivery
2017 - 2019
Delivered during Post-Doctoral Fellowship at IIT Hyderabad
Successfully delivered a DRDO (Defense Research Development Organization, Ministry of defense, Govt of India) project where the primary task was to design a compressed version of VGGNET for object classification and tracking.
Academic Experience
Senior Research Officer (Post Doctoral Research Associate PDRA)
August 2023 - Present
School of Computer Science and Electronic Engineering, University of Essex, UK
Senior Research Associate
May 2022 - August 2023
Indian Institute of Technology (IIT), Hyderabad, India
National Post-Doctoral Fellow (DST, SERB, Govt of India)
August 2017 - August 2019
Department of Electrical Engineering, Indian Institute of Technology (IIT), Hyderabad, India
Newton Bhabha Fellow
2015 - 2016
University of Southampton, Hampshire, England
Efficient algorithm design and prototype hardware architecture for functional brain connectivity formulation from multi-channel EEG data, extracting markers for quantitative characterization, and computing the temporal variability of such markers to convey the implicit nature of cognitive efficiency.
Assistant Professor (Teaching)
July 2010 - January 2012
Sikkim Manipal Institute of Technology, Sikkim, India
Held a one-and-a-half-year teaching position. Supervised various final year graduate engineering projects. Following this role, joined full-time research under the DST Inspire scheme, Govt. of India, where I continued to supervise masters projects and was invited to workshops and seminars for training and speaking engagements.