jetson-inference VS grim

Jetson Nano’s are obsolete and no longer supported; but to answer your question, this might be a good place to start.

I’m trying to do the tutorial, where they retrain the neural network to detect fruits (jetson-inference/pytorch-ssd.md at master · dusty-nv/jetson-inference · GitHub 1)

Jetson-Inference is another amazing resource to get started on. This will allow you to try out a number of neural networks (classification, detection, and segmentation) all with your own data or with sample images included in the repo.

Hello! I'm looking for some pre trained image classification models to use on a Jetson Nano. I already know about the model zoo and the pre trained models included in the https://github.com/dusty-nv/jetson-inference repo. For demonstration purposes, however, I need a model trained on small objects from the context of production, ideally nuts, bolts, and similar small objects. Does anyone happen to know a source for this? Thanks a lot!

> They provide some SSD-Mobilenet-v2 here: https://github.com/dusty-nv/jetson-inference

I was aware of that repository but from taking a cursory look at it I had thought dusty was just converting models from PyTorch to TensorRT, like here[0, 1]. Am I missing something?

> I get 140 fps on a Xavier NX

That really is impressive. Holy shit.

[0]: https://github.com/dusty-nv/jetson-inference/blob/master/doc...

[1]: https://github.com/dusty-nv/jetson-inference/issues/896#issu...

If you have a screen and keyboard and mouse plugged into the Nano, I would recommend starting with Hello AI World on https://github.com/dusty-nv/jetson-inference#hello-ai-world

Well, don't get me wrong. I do like my Jetson Nano. For a hobbyist who likes to tinker with machine learning in their spare time it's definitely a product cool and there are quite a few repositories on Github[0, 1] with sample code.

Unfortunately… that's about it. There is little documentation about

- how to build a custom OS image (necessary if you're thinking about using Jetson as part of your own product, i.e. a large-scale deployment). What proprietary drivers and libraries do I need to install? Nvidia basically says, here's a Ubuntu image with the usual GUI, complete driver stack and everything – take it or leave it. Unfortunately, the GUI alone is eating up a lot of the precious CPU and GPU resources, so using that OS image is no option.

- how deployment works on production modules (as opposed to the non-production module in the Developer Kit)

- what production modules are available in the first place ("Please refer to our partners")

- what wifi dongles are compatible (the most recent Jetson Nano comes w/o wifi)

- how to convert your custom models to TensorRT, what you need to pay attention to etc. (The official docs basically say: Have a look at the following nondescript sample code. Good luck.)

- … (I'm sure I'm forgetting many other things that I've struggled with over the past months)

Anyway. It's not that this information isn't out there somewhere in some blog post, some Github repo or some thread on the Nvidia forums[2]. (Though I have yet to find a reliably working wifi dongle…) But it usually takes you days orweeks to find it. From a product which is supposed to be industry-grade I would have expected more.

[0]: https://github.com/dusty-nv/jetson-inference

https://github.com/dusty-nv/jetson-inference#system-setup

"GNOME doesn't support wlr-screencopy-unstable-v1, which is the protocol grim uses to take screenshots." Source.

It's in the README

there is grim which is supposed to work under Wayland but it seems like it only works under swayWM, the reason why i need a cli tool is that I want to build Rofi script on top of it, I'm a ware spectacle and it is a very great option but I would prefer a Rofi based tool .

sway laptop user here (for almost 2 years I think?).

I spent a little while on this, but I migrated from i3, so I just ported every little section of my config bit by bit.

In terms of battery bar and other "bar" type things, I use waybar[0] which basically does all the things you'd expect by default (just install and it "works").

For multi-monitor, config, I initially setup with wdisplays[1] (think arandr for wayland) and then manually copied the positions into my sway config. Monitor positioning was the only thing I needed to setup (and telling it that one monitor was HDPI) and then all of the scaling and everything worked perfectly. This was my biggest selling point for wayland, I now get nice crisp fonts and application scaling works nicely (which was not the case with X).

volume control from the keyboard took no time, just a couple of extra lines.

There was some stuff to do with the clipboard (wl-clipboard[2]) and screenshots (grim[3] + slurp[4]) that required some setup, but again, just a few lines, and didn't take much mental load.

Oh and I needed to change my notifications daemon(dunst[5]), and chose to change my program launcher to one with a nicer interface and cleaner fonts (wofi[6]).

I think that's all the tweaking that I did. Oh, and I needed to do something with pipewire to sort out screensharing at the start, don't remember that too well though...

[0] https://github.com/Alexays/Waybar

[1] https://github.com/artizirk/wdisplays

[2] https://github.com/bugaevc/wl-clipboard

[3] https://github.com/emersion/grim

[4] https://github.com/emersion/slurp

[5] https://github.com/dunst-project/dunst

[6] https://hg.sr.ht/~scoopta/wofi

grim might work well but you'll probably have to write a shell script or something to keep track of the user preferences. You'll also need slurp if you want to select a region to screenshot.

Screenshots: grim + slurp + swappy

Since you have already noticed that it does not work with Wayland, it is a strange requirement to take a screenshot "with it". Why not use a tool that does work with wayland, e.g. Grim?

# Details I'm an English speaker living abroad, and while I'm trying to learn the local language it's real hard. I found myself popping open a browser to use deepl quite frequently, or trying to find translator plugins for several different applications. To make this process easier, I wrote a script (bound to a hotkey) which will screenshot a selected area, OCR it, translate it to english, and show a notification with the translated text. It also copies the translated text to the clipboard. Why screenshot + OCR rather than just selecting and copying text? Images and screen-sharing, mostly. I think this is just a really cool way to show how the hard parts have usually been done for you, and all you need to do is put the blocks together. ## Implementation I'm running sway, so the several of the tools are Wayland specific. You could easily swap them out for xorg compatible variants if you like. The script is [here](https://github.com/rbuchberger/dotfiles/blob/master/scripts/screenshot\_translate). The toolchain is: * [Slurp](https://github.com/emersion/slurp) - select an area * [Grim](https://github.com/emersion/grim) - screenshot that area * [Tesseract](https://github.com/tesseract-ocr/tesseract) - OCR * [Translate Shell](https://www.soimort.org/translate-shell/) - Translation CLI * [Mako](https://github.com/emersion/mako) - Notification window Mako needed a little configuration to show long form text: [category=translation] width=900 height=1200 Edit: added details and links for the tools used.

I don't wanna poop on their parade, but haven't Wayland screenshotters been around for a while? https://github.com/emersion/grim

That one has at least been around for long enough, and has worked perfectly under Sway for long enough, that I had to look up its name because I had it bound to a hotkey and had forgotten what it was called.