- Currently, I am trying to devise a new algorithm for classifying simple 3D objects given meshes as input such that changes in position, size, thickness, rotation and some degree of noise can be dealt with.
- To solve this problem, I have come up with a new 3D filter. At present, I am trying to refine the same.
- Future Plans: The aim is to recognize the class of the given object. Get the best grasping point and associated grasp from a database of precomputed grasps for each class of object. Scale the grasping point according to the size of the object, and account for the position and orientation of the given object to determine the 3D location of the grasping point.
- Making preparations to start a fun project on the side: To enable our iCub to throw balls in a bin. (Every Friday)
- Gave a presentation of my work in VisLab on October 12, 2010
- Attended IURS 2010 in Benicassim, Spain