Sujal Vijayaraghavan
Research Subjects
Research Interests
Sujal Vijayaraghavan has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Rajiv Dubey
University of South Florida
2 shared publications - Redwan Alqasemi
University of South Florida
2 shared publications - Sudeep Sarkar
University of South Florida
2 shared publications
Latest Publications
- Open Access
LocaliseBot: Multi-view 3D object localisation with differentiable rendering for robot grasping
Robot grasp typically follows five stages: object detection, object localisation, object pose estimation, grasp pose estimation, and grasp planning. We focus on object pose estimation. Our approach relies on three pieces of information: multiple views of the object, the camera's extrinsic parameters at those viewpoints, and 3D CAD models of objects. The first step involves a standard deep learning backbone (FCN ResNet) to estimate the object label, semantic segmentation, and a coarse estimate of the object pose with respect to the camera. Our novelty is using a refinement module that starts from the coarse pose estimate and refines it by optimisation through differentiable rendering. This is a purely vision-based approach that avoids the need for other information such as point cloud or depth images. We evaluate our object pose estimation approach on the ShapeNet dataset and show improvements over the state of the art. We also show that the estimated object pose results in 99.65% grasp accuracy with the ground truth grasp candidates on the Object Clutter Indoor Dataset (OCID) Grasp dataset, as computed using standard practice.
Open AccessLocaliseBot: Multi-view 3D object localisation with differentiable rendering for robot grasping
Robot grasp typically follows five stages: object detection, object localisation, object pose estimation, grasp pose estimation, and grasp planning. We focus on object pose estimation. Our approach relies on three pieces of information: multiple views of the object, the camera's extrinsic parameters at those viewpoints, and 3D CAD models of objects. The first step involves a standard deep learning backbone (FCN ResNet) to estimate the object label, semantic segmentation, and a coarse estimate of the object pose with respect to the camera. Our novelty is using a refinement module that starts from the coarse pose estimate and refines it by optimisation through differentiable rendering. This is a purely vision-based approach that avoids the need for other information such as point cloud or depth images. We evaluate our object pose estimation approach on the ShapeNet dataset and show improvements over the state of the art. We also show that the estimated object pose results in 99.65% grasp accuracy with the ground truth grasp candidates on the Object Clutter Indoor Dataset (OCID) Grasp dataset, as computed using standard practice.
- 19 February 2023
- Published by Springer Nature
- p. 699-711
LocaliseBot: Multi-view 3D Object Localisation with Differentiable Rendering for Robot Grasping
- 19 February 2023
- Published by Springer Nature
- p. 699-711
LocaliseBot: Multi-view 3D Object Localisation with Differentiable Rendering for Robot Grasping