I dont have the cash to spare for the tx2 but i will at some point.
would like to start building out some ideas.
could i stream in the video from somewhere else like a locally networked android phone?
or like a hololens for example?
maybe even a webcam? can it process multiple streams of video? for example if i had 10 cameras not locally connected to the device actually running the jetpack software?
While JetPack is intended for setting up Jetson from host PC, the software components included with JetPack (like CUDA toolkit, cuDNN, TensorRT, ect.) all have PC versions available too. So you will just want to install those on your machine. It is unclear if they will work with GPU acceleration from within VM on Windows. You would likely be better off by dual-booting your Windows machine with an Ubuntu 16.04 install on the side.
The webcam, if you use V4L2, that has the same API between PC and Jetson.
thanks!
so i can run all the software on my PC. ok got that.
but can i run the the webcam V4L2 from a raspberry pi and stream the video to the PC?
what if i have an old linux laptop with old embedded webcam? will that let me stream to the softwares running on the windows PC?
Hi juan.suero,
We have usecases of udp and rtsp:
[url]https://devtalk.nvidia.com/default/topic/1027423/jetson-tx2/gstreamer-issue-on-tx2/post/5225972/#5225972[/url]
[url]https://devtalk.nvidia.com/default/topic/1018689/jetson-tx2/vlc-playing-gstreamer-flow/post/5187270/#5187270[/url]
[url]https://devtalk.nvidia.com/default/topic/1014789/jetson-tx1/-the-cpu-usage-cannot-down-use-cuda-decode-/post/5188538/#5188538[/url]
whats the difference between the tx2 and the tx1 other than processing power?
I dont mind using low power one but i want to use the latest software.
the nvidia jetpack comes for both?
what about all the deep learning tools? is there even one API missing when i use the tx1 over the tx2?
the language here is kinda fuzzy sales language where its seems to be the same.
Hi juan,
r28.1 is for both tx1 and Tx2 https://developer.nvidia.com/embedded/dlc/l4t-documentation-28-1
r28.1 is kernel 4.4. You can use Jetpack 3.2 to flash both.
For HW deviation between Tx1 and TX2, you may check oem design guide:
http://developer.nvidia.com/embedded/dlc/jetson-tx1-oem-product-design-guide
https://developer.nvidia.com/embedded/dlc/jetson-tx2-oem-product-designguide
thanks but too late i already ordered the tx2. instead of using the onboard cameras can i stream the video from a source on the local lan?
can i also do object positioning? i dont mean the position of the tx2 i mean the position of where the video was taken