Overview
The deep learning and AI research area comprises multiple teams that work on a variety of research topics, all with a clear focus on solving real-world research problems. These teams are currently exploring program synthesis, reinforcement learning for robotic control, and teaching agents to negotiate with humans.
Fields
Some focus areas include:
Sensor analytics: This group employs deep learning for anomaly detection and machine health estimation from time series sensor data using its iSense platform
WildML: This group is working on leveraging human elastic labor via crowdsourcing to enhance deep learning solutions through its Samuhaa platform. This involves research on crowd reputation management and crowd consensus algorithms, as well as deep models for speech, vision, and text
Deep vision: This team resolves a variety of vision problems involving data extraction from images, as well as medical image processing using deep learning techniques
Augmented reality (AR): This team works on a number of vision and graphics problems, ranging from gesture recognition, model compression, and saliency detection, to multi-label placement, semantic segmentation, and object recognition
People & Patents
Research team: Led by Senior Scientist Lovekesh Vig, the team includes Pankaj Malhotra; Shirish Karande; Monika Sharma; and Ramya Hebbalaguppe
Academic partners: BITS Goa, India; and IIT Delhi, India
Patents and publications (2016 onwards): This team has published their work at different conferences and journals