About

Umar Asif is a leading researcher in robotics and artificial intelligence, with key contributions spanning multi-legged locomotion, robotic grasping, and object recognition. His early work on hexapod robots, including a fault-tolerant adaptive gait for navigating unstructured terrains and a balance stabilization method using ZMP-based pattern generation, established foundational techniques for robust locomotion—earning 377 citations for his 2012 paper. Asif’s impact extends to deep learning for robotics, where he developed GraspNet (167 citations), an efficient CNN architecture enabling real-time grasp detection on low-powered devices, and the Densely Supervised Grasp Detector (DSGD), which fuses multi-level features for high-confidence grasp prediction. His hierarchical cascaded forests framework for RGB-D object recognition and grasp detection (157 citations) further demonstrates his ability to integrate perception and manipulation. With over 800 total citations, Asif’s work bridges classical control and modern AI, advancing practical robotic systems for challenging environments. His achievements include pioneering efficient models that balance accuracy with computational constraints, making him a key figure in deployable robotic intelligence.

Research Focus

Key Achievements

9
H-Index
15
Papers
860
Total Citations
57
Avg Citations/Paper
🏆 Most Cited Paper
On the Improvement of Multi-Legged Locomotion over Difficult Terrains Using a Balance Stabilization Method
377 citations · 2012
📈 Most Prolific Year: 2011 (6 Papers)
🤝 Key Collaborators: 9
🏛 Institutions: University of the Sciences, IBM Research - Australia, National University of Sciences and Technology, The University of Western Australia

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago