$ cd ros2_ws/src/
$ git clone <repo name>

- torch2trt related model file updates <br/>
    - Clone the MiDaS repository<br/>
    

$ cd ros2_ws/src/ros2_monocular_depth/ros2_monocular_depth
$ git clone https://github.com/intel-isl/MiDaS

    - To resolve a torch2trt tracing error, modify the following line https://github.com/intel-isl/MiDaS/blob/f275ca1c6f9af17aabe6f1e024b2084d7b84abb0/midas/blocks.py#L117 to read<br/>
    

x = nn.functional.interpolate(
x, scale_factor=self.scale_factor, mode=self.mode, align_corners=False
)

    - Download the MiDaS small model weights into `model_dir`. Name them midas_trt.pth.<br/>

- Build and install `ros2_monocular_depth` package <br/>

## Run

- Change Power Mode for Jetson

``` sudo nvpmodel -m2 
``` (for Jetson Xavier NX) <br/>

``` sudo nvpmodel -m0 
``` (for Jetson Xavier and Jetson Nano) <br/>
- Input Images are captured using 
```image_tools
``` package <br/>

``` ros2 run image_tools cam2image 

$ ros2 run ros2_monocular_depth depth_estimation --ros-args -p model_dir:='<absolute-path-to-base_dir>'