Grasping with Cognitive Humanoid Robots using Deep learning

This project is based on the development of a synthetic dataset for grasping basic primitive shapes for cognitive humanoid robots in a simulated environment.  * The system will be first trained on a deep learning framework with the annotated dataset with ground truth values for the robo

2025-06-28 16:32:46 - Adil Khan

Project Title

Grasping with Cognitive Humanoid Robots using Deep learning

Project Area of Specialization RoboticsProject Summary

This project is based on the development of a synthetic dataset for grasping basic primitive shapes for cognitive humanoid robots in a simulated environment. 

* The system will be first trained on a deep learning framework with the annotated dataset with ground truth values for the robotic arm and hand. 

* The results generated from the system could be employed on  a humanoid robotic hand for various purposes, i.e robotic surgeries, industries, house keeping etc.

Project Objectives

development of synthetic dataset for basic primitive shapes in iCub simulated environment.

* generation of various position, orientation and kinematic chain data from the iCub robotic arm for object grasping.

* training a deep learning framework using transfer learning with the dataset of objects to estimate joint angles.

* The comparative analysis with other deep learning frameworks can show robustness of the networks.

Project Implementation Method

Various objects in the iCub simulator will be loaded in front of a table.

* The iCub robotic arm will be used to grab those objects.

* The information related with the arm i.e kinematic chain, position and orientation of the hand will be recorded.

* A dataset with information of the objects will be annotated with kinematic chain and robotic hand.

* A deep learning framework for example AlexNet, will be trained with these information.

Benefits of the Project

The proposed system will provide the ability for humanoid robotic arms for grasping various objects.

* It could be further used in medical sciences i.e robotic surgery, industries, house keeping etc.

Technical Details of Final Deliverable

 humanoid robotic hand grassping for various purposes, i.e robotic surgeries, industries, house keeping etc.

Final Deliverable of the Project HW/SW integrated systemType of Industry IT , Security Technologies Artificial Intelligence(AI), RoboticsSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources
Item Name Type No. of Units Per Unit Cost (in Rs) Total (in Rs)
Total in (Rs) 69500
Nvidia Geforce 1080 GPU Equipment16950069500

More Posts