Shariq Iqbal
Papers
1
Total Citations
2
H-Index
1
About
Shariq Iqbal is a researcher advancing the frontiers of robotic manipulation through deep reinforcement learning and sim-to-real transfer. His work centers on enabling robots to perform semantically informed, directional grasping—a critical capability for real-world applications where a robot must not only identify an object but grasp it from a specific, task-relevant orientation. In his notable 2020 paper, "Toward Sim-to-Real Directional Semantic Grasping," Iqbal tackles this challenge by training a double deep Q-network (DDQN) to map low-resolution RGB images from a wrist-mounted camera directly to Q-values, allowing the robot to learn grasping policies entirely in simulation before transferring them to physical hardware. This approach bridges the reality gap, demonstrating that complex, directional manipulation skills can be learned efficiently without extensive real-world data. While his citation count is still growing, Iqbal’s contributions are foundational for next-generation robotic systems that require precise, context-aware interaction with their environment. His work sits at the intersection of computer vision, reinforcement learning, and robotics, offering a scalable pathway toward more autonomous and capable robotic assistants.
Research Focus
Key Achievements
Top Papers
- 1Toward Sim-to-Real Directional Semantic Grasping2 citations · 2020