Riad Akrour

Latest

• Embodied Active Learning under Limited Annotation and Navigation Budget for Object Detection
• Compatible natural gradient policy search
• Projections for Approximate Policy Iteration Algorithms