Joydeep Biswas

Robots can better adapt to new terrains by predicting operator preferences.

Using game theory, robots coordinate movements in narrow areas to avoid collisions.

This dataset aids robots in better understanding urban environments.

Robots adapt to diverse terrains using self-collected data.

ObVi-SLAM improves robot localization by combining visual features and object detection.

A framework for better understanding how people engage with robots using video simulations.

A framework for machines to learn user preferences from visual data.

Dynamic Model Predictive Shielding enhances safety and efficiency in reinforcement learning.

Imitation learning helps robots move better in busy environments.

A new framework enhances smaller models' abilities in robot programming.

A new method helps robots see their surroundings clearly without human input.

CLOVER enhances object recognition in robots using context and innovative learning methods.

Robots can now answer questions using long-term memory.

Robots learn to interact with everyday articulated objects for home assistance.

New robot designs aim to assist those with vision impairments.

SPOT learns tasks by watching humans, making robots more helpful.

New method enhances how robots locate themselves on the ISS.