Which are the top platforms or libraries to use to do computer vision (eg object detection, image classification) in 2024?
I know pytorch, tensorflow, huggingface, etc. have models, but there’s so much out there I thought I’d try to get some opinions from peers.
As far as platforms are concerned, try the ones from google or any other cloud provider.
I long suffered from research-itist and only overcame it by ripping off the bandaid and just-doing-it.
You can easily spin up a model and do some inference and get the outcome by having the right tools selected…
One thing that literally nobody willingly talks about up front is the reality of cost when they recommend a, b, or c tools. Specifically computer vision platforms from the top tier providers (google, aws, azure, etc).
This missing piece will either break the bank or end up blocking you during your discovery phase.
If you don’t have resources at your disposal or aren’t an ops/infra guy check out lightning ai or similar “studio” development environments. You can spin up/change/test out a huge selection of specific GPU models, click to test, and so on.