jasongill

If you, like me, just got a new Nvidia Jetson Orin Nano Super Developer Kit and are struggling to get the examples provided by Nvidia to run due to the following error:

ImportError: libnvdla_compiler.so: cannot open shared object file: No such file or directory

then you probably don't want to have to deal with downgrading the flashed version of things just to see if it will work. The issue is that the libnvdla_compiler.so file is not included in the latest (36.4.2 as of this writing) version of the Jetpack software which you're instructed to install to upgrade the firmware on the device.

The fix is to grab the file from the .deb file from the previous version. Running the following command will fix the issue:

MonocleSecurity

#!/bin/bash
set -e
##### CHECK PARAMETERS #####
PRODUCTION=0
while [[ "$#" -gt 0 ]]; do
    case $1 in
        -p|--production) PRODUCTION=1 ;;
        *) echo "Unknown parameter passed: $1"; exit 1 ;;
    esac
    shift

discrimy

#!/bin/env bash
set -e

# USAGE:
# COPY .devcontainer/ /app/.devcontainer/
# ARG VSCODE_VERSION=2ccd690cbff1569e4a83d7c43d45101f817401dc
# RUN bash .devcontainer/preinstall-vscode-extensions.sh $VSCODE_VERSION /app/.devcontainer/devcontainer.json
# 
# Preinstall vscode server extensions.
# Uses vscode of provided version (git commit hash) to download extensions inside container

unitycoder

NeRF

Repo: https://github.com/NVlabs/instant-ngp

Steps (see their readme also)

• install CUDA toolkit https://developer.nvidia.com/cuda-downloads?target_os=Windows&target_arch=x86_64&target_version=10&target_type=exe_local
  • you can remove all the extras, like geforce experience and docs etc *i installed 11.4
• install cmake https://cmake.org/download/ (latest release, not RC), i took 3.22.5 *3.23 doesnt work?
• install optional (for DLSS) https://vulkan.lunarg.com/sdk/home#windows (base sdk only)
• make folder somewhere, lets say E:\apps\NeRF\

utkonos

import io
import pathlib

import pycdlib

ubuntu = pathlib.Path('ubuntu-22.04.1-live-server-amd64.iso')
new_iso_path = pathlib.Path('ubuntu-22.04.1-live-server-amd64-auto.iso')

iso = pycdlib.PyCdlib()
iso.open(ubuntu)

faisalthaheem

#from rtsp server android app
'rtspsrc location=rtsp://192.168.1.193:5554/camera latency=0 ! rtph264depay ! avdec_h264 ! videorate max-rate=1 ! decodebin ! videoconvert ! video/x-raw, format=BGR ! queue max-size-buffers=10 ! jpegenc quality=25 ! jpegparse ! appsink name=sink'

#big buck bunny
'rtspsrc location=rtsp://184.72.239.149/vod/mp4:BigBuckBunny_115k.mov latency=0 ! rtph264depay ! avdec_h264 ! videorate max-rate=15 ! decodebin ! queue max-size-buffers=10 ! jpegenc quality=25 ! jpegparse ! appsink name=sink'

# example of mjpeg stream from a typical axis camera or esp32
# souphttpsrc location=http://x.y.z.a:8090/mjpg/video.mjpg?timestamp=1564650227443 ! multipartdemux single-stream=true ! image/jpeg,width=320,height=240,framerate=3/1 ! appsink name=sink

#Example of an RTSP source

martin-niklasson

The background is that the default OpenWRT image for Raspberry Pi lacks drivers for the LAN port (RTL8111) on the DFRobot board. DFRobot provide an image, but that one is made from a quite old daily snapshot of OpenWRT which is not entirely stable. Also, it is better to know how to fish than be given a fish.

DFRobot's page about the router board:
https://wiki.dfrobot.com/Compute_Module_4_IoT_Router_Board_Mini_SKU_DFR0767

OpenWRT main page:
https://openwrt.org/

You are expected to have some general knowledge on Linux and OpenWRT. The instructions for flashing the resulting firmware image are presented on DFRobot's page.

nilesh0109

import cv2
import numpy as np

def get_inverse_pespective(perspective_matrix: np.array)-> np.array:
  """
  This method calculates the inverse of prespective matrix by homography. 
  - Takes 4 random points on the floor plane(destination_plane) and calculates the corresponding points 
  on the camera image plane(src_plane) using perspective matrix.
  - Calculates the Homography matrix to map any point in image plane to floor plane.
  Parameters

jackersson

#!/bin/bash
RECDIR=/mnt/record/minisio/$(date +%Y-%m-%d)
RECTIME=$(date +-%H:%M:%S)
if [ ! -d "$RECDIR" ]; then
mkdir -p $RECDIR
fi

cd $RECDIR

for i in {0..6}; do

gumdropsteve

# folder where the image is already saved (without bounding boxes)
data_dir = 'media/stable_numbers/number_crops/'

# folder where the new image with bounding boxes should be saved
out_dir = 'media/copy_labeled_number_crops/'

# go through the unique file names (because files can have more than 1 bounding box)
for file in new_df.filename.unique():
    
    # load in the image and .loc[] the rows where it's referenced (each row represents 1 bounding box)