Comment

Comment by Ftuuky

I'm trying to learn something about deep learning and do an end-to-end project with computer vision.

I have a raspberry pi and picamera and wanted to detect the pigeons in my balcony and then play a sound or something to shoo them away.

But it's going nowhere, I'm too dumb to even start properly :(

- Nvidia and CUDA stuff is so hard, I can't set it up properly no matter what

- Tried YOLO but without CUDA and OpenCV I can't run it in video. Don't know how to fix it

- Tried to copy other projects but can't find anything that I can parse with my amateur brain. I get lost and doesn't matter how many youtube videos I watch or stackoverflow pages I check, it's errors after errors after errors.

- Tried in windows but that's not viable. Installing Ubuntu nearly broke my pc and somehow a virtualbox messes up the whole thing. Currently looking at this.

So yeah big mess, I'm way over my head and it's not fun anymore. But I still want to shoo away the pigeons and love the idea of learning more about DL/CV but guess I need to learn about the basics first, practice in other things before doing this.

parent Ftuuky 6 years ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton