Mahsa Baktashmotlagh

Queensland University of Technology, The University of Queensland

Papers

Total Citations

H-Index

About

Dr. Mahsa Baktashmotlagh is a leading researcher in computer vision and machine learning, with a focus on enabling intelligent systems to understand and anticipate human actions and natural language. Her work bridges the gap between real-time perception and robotic interaction. In her highly cited 2017 paper, "On Encoding Temporal Evolution for Real-time Action Prediction," she pioneered methods for anticipating motion evolution in video frames, a critical capability for autonomous cars and robots that must predict future actions in real time—a contribution that has garnered 5 citations and influenced the field of action anticipation. She further advanced human-robot communication in her 2019 work, "Object Graph Networks for Spatial Language Grounding," which tackles the challenge of enabling robots to interpret spatial references like "the cup nearest to the plate." By developing graph-based models that map natural language phrases to visual scenes, Dr. Baktashmotlagh has made significant strides in spatial language grounding, with 2 citations reflecting its niche impact. Her research is notable for its practical applications in domestic robotics and autonomous systems, demonstrating a commitment to making AI more intuitive and responsive to human needs.