Dr. Ashish Gupta

 Leading a ML team in developing R&D solutions for B2B and B2C customers.  Developing Deep Learning based solutions for visual detection, classification, and image generation for consumer appliances using CNN, Vision Transformer, YOLOv8, Generative Adversarial Network, and Stable Diffusion.  Growing company’s capability in Generative AI and Multi-Modal Large Language Models for data generation, captioning, and VQA, using Vertex AI, PALM, Gemini.

R&D in Deep Learning, Computer Vision, and Signal Processing. Leading team of PhD research scientists and engineers. Management and development of Machine Learning (ML) solutions for Nanotechnology research. • Research and develop end-to-end predictive data analytics for scanning probe microscopy data. • Collaborate with Data Engineering team to build an ML infrastructure for automation. • Inspiring a culture shift towards a data-driven mindset.

Data Scientist in the Data and Analytics team at Geotab. My focus is exploring and developing Machine Learning and Data Science solutions. R&D into analysis of video for pedestrian and vehicular activity analytics, particularly at road intersections. My current focus is incorporating state-of-art visual detectors (FRCNN, YOLO v4) into our pipeline. Collaborating with University of Kingston. Evaluating potential business partners for tasks like data annotation and deep learning training automation.

Research: Collaborated in cross-disciplinary research on the application of Deep Learning to Medical Image Analysis. Developed research project and software (using PyTorch) for automated assessment of quality in structural MRI data using deep learning (DNN, ResNet, GAN). Delivered seminars to the CCIPD center on feature representation and deep learning. Mentored graduate students and visiting scholars. Team lead for project on exploratory analysis of machine learning for autism detection in MRI data.

Research on image and video forensic analysis, as part of a large consortium of more than thirty institutions, for ‘MediFor’ project funded by DARPA. I focus on developing machine learning and computer vision methods to analyze media content for malicious tampering and also generate ‘smart’ manipulations to aid development of counter-forensic methods. I work in a team at RIT on Deep-Learning based methods. I collaborate with external partners including PAR technologies, who liaise between DARPA and RIT.

Research in computer vision and machine learning. The focus is localization of camera and 3D trajectory estimation. The data will be acquired and processed on a mobile platform.

Research and development of novel image aesthetics and interestingness descriptors. The objective is automatic editing of large corpus of photographs based on perceptual metrics of quality.

Teaching assistant and lab demonstrator for undergraduate engineering students at Surrey.

Research and Development in the Applied Vision Group. Algorithm and software solutions for advanced computer vision industrial applications

Computer Vision, Machine Learning

Signal Processing, Computer Vision, Robotics

Control engineering, Signal and Image Processing, Instrumentation engineering