Papers
57
Total Citations
840
H-Index
17
About
Hamidreza Kasaei is a robotics researcher whose work spans 3D object recognition, robot manipulation, lifelong learning, and autonomous navigation. His research addresses one of the field's most persistent challenges: enabling robots to perceive, learn about, and interact with objects in open-ended, real-world environments. Early contributions, including an interactive open-ended learning framework (2015, 42 citations) and a perceptual memory system for grounding semantic representations (2014, 32 citations), established him as a pioneer in adaptive robot perception. His GOOD descriptor (2016, 58 citations) became a widely adopted tool for 3D object recognition and manipulation, while OrthographicNet extended this work through deep transfer learning. Kasaei further advanced robot capability through continuous curriculum learning for reinforcement-based reaching tasks (2020, 52 citations) and multi-view grasping in cluttered environments (2022, 36 citations). More recently, his integration of large language models into visual target navigation — L3MVN (2023, 81 citations) — has drawn significant attention, reflecting his forward-looking focus on combining common-sense AI with autonomous robotics. Collectively, his work represents a coherent vision of robots that learn continuously, reason intelligently, and act reliably alongside humans.
Research Focus
Key Achievements
Top Papers
- 1L3MVN: Leveraging Large Language Models for Visual Target Navigation81 citations · 2023
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9The RACE Project25 citations · 2014
- 10Frontier Semantic Exploration for Visual Target Navigation23 citations · 2023