Sulabh Kumra
Sulabh Kumra
Good point, the gaussian filter is used to mostly filter out the quality function. It probably shouldn't be applied to the angle. Please feel free to open a pull request...
The runtime speed is a function of the hardware being used for computation. The results reported in the paper are for the hardware we used for the experiments. You can...
You should be able to run evaluate.py on your own model. Here is an example: `python evaluate.py --network --dataset cornell --dataset-path --iou-eval `
Apologies for late response. This is just poor naming of the variables. Please feel free to open a pull request to refactor it.
You can extract the Cornell dataset to a directory and it should have data folders 1-10 from the dataset. The path to this directory can be specified using `--dataset-path `.
Yeah, probably not the best implementation. Please feel free to open a pull request to refactor it.
The code in this repository is currently not compatible with ROS.
The information in `camera_pose.txt` and `camera_depth_scale.txt` depends on your hardware. `camera_pose.txt` contains the pose of the camera in the robot frame. `camera_depth_scale.txt` contains the depth scale of the camera. If...
You can use the `ds-shuffle` parameter for it. The default is object-wise (OW) and setting it to `True` with give you image-wise (IW).
Are you using a ROS node for camera? The current code doesn't support ROS and that's why it is unable to connect to the camera/device.