https://github.com/um-arm-lab/contact_shape_completion

Shape completion using contact measurements. Successor of PSSNet
https://github.com/um-arm-lab/contact_shape_completion

Last synced: 4 days ago
JSON representation

Shape completion using contact measurements. Successor of PSSNet

• Host: GitHub
• URL: https://github.com/um-arm-lab/contact_shape_completion
• Owner: UM-ARM-Lab
• Created: 2020-11-10T15:18:22.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2021-10-24T23:05:24.000Z (about 3 years ago)
• Last Synced: 2023-08-24T16:14:16.330Z (about 1 year ago)
• Language: Python
• Size: 6.58 MB
• Stars: 0
• Watchers: 10
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# CLASP

Constrained LAtent Shape Projection (CLASP) combines a shape completion neural network with contact measurements from a robot. This repo also contains the [full paper](https://github.com/UM-ARM-Lab/contact_shape_completion/blob/main/CLASP_full.pdf) and [additional plots](https://github.com/UM-ARM-Lab/contact_shape_completion/blob/main/CLASP_additional_details.pdf).

# Quick Start

1. Set up ROS

2. Clone this repo in your ROS path. Rebuild (e.g. catkin build), re-source

3. Install dependencies

4. Download datasets and pretrained models by running `shape_completion_training/scripts/download_pretrained.py`

### Data Analysis

Trial results are in `./evaluations`

To recreate the plots from the paper using the pre-run trials, `contact_completion_evaluation.py --plot`

### Rerun shape completion experiments and visualize performance

To rerun shape completion experiments using the robot motion and contacts as recorded from the paper, in separate terminals run:

1. `roslaunch shape_completion_visualization live_shape_completion.launch` (Rviz will start)

2. `roslaunch shape_completion_visualization transforms.launch` (Sets transforms between robot and shape objects)

3. `contact_completion_evaluation.py --plot --regenerate`  (You will see many shape being generated and updates. Will take over an hour to complete)

# Full Stack

The full experimental setup requires running a simulated, or real robot, which moves and contacts objects.

To build the software stack used in the experiments, set up the dependencies.

Detailed instructions are in the subfolder [https://github.com/UM-ARM-Lab/contact_shape_completion/tree/main/contact_shape_completion]

Then run

1. `roslaunch shape_completion_visualization live_shape_completion.launch` (Rviz will start)

2. Launch the robot stack (see detailed instructions in subfolder)

3. `store_simulation_examples --trial [PRETRAINED_NETWORK_NAME] --scene [SCENE_NAME] --store`