Most active commenters
  • roryclear(5)
  • diggan(3)

←back to thread

Show HN: Clearcam – Add AI object detection to your IP CCTV cameras

(github.com)
230 points roryclear | 17 comments | 24 Aug 25 11:34 UTC | HN request time: 0.704s | source | bottom

This runs YOLOv8 + bytetrack with Tinygrad detections (depending on user config) are saved and can be sent to the companion iOS app along with a notification, all video processing is done locally, all footage is encrypted before leaving your computer, and the sending notifications + videos part is optional. This uses tinygrad, so it runs well on my apple silicon macs and should be able to run on a lot of hardware (or will be able to when I remove other deps).
1. snickerdoodle12 ◴[24 Aug 25 13:44 UTC] No.45004203[source]▶
how does this compare with frigate?
replies(3): >>45004242 #>>45004269 #>>45005215 #
2. roryclear ◴[24 Aug 25 13:50 UTC] No.45004242[source]▶
fewer features, easier setup, with more GPUs supported. (I've not used frigate myself though, only watched videos)
replies(2): >>45004287 #>>45009819 #
3. diggan ◴[24 Aug 25 13:54 UTC] No.45004269[source]▶
User of Frigate here. Seems these are some pretty major differences of what you can do for free with Frigate, but if you use Clearcam, you need to pay for "Clearcam Premium":

- View your live camera feeds remotely.

- Receive notifications on events (objects/people detected).

- View event clips remotely.

- End-to-end encryption on all data.

What neither of the solutions seem to have, is encryption at rest. But I guess others, just like me, rather encrypt the volume/storage itself, instead of leaving it up to applications anyways, so might/might not matter for you.

replies(1): >>45005848 #
4. diggan ◴[24 Aug 25 13:56 UTC] No.45004287[source]▶
Where can I find the list of supported GPUs? Frigate been able to handle everything I've tried so far, all from Nvidia and AMD GPUs to even Intel iGPUs.
replies(2): >>45004334 #>>45004471 #
5. d0ugal ◴[24 Aug 25 14:05 UTC] No.45004334{3}[source]▶
I have used Frigate for years, I think early on it didn't support all of those GPUs. So it might be that said videos are out of date.
replies(1): >>45004615 #
6. serf ◴[24 Aug 25 14:23 UTC] No.45004471{3}[source]▶
same here -- it's also among one of the only things to support Coral devices and RPi video cores.

I would imagine any GPGPU compute-capable pre-CUDA thing probably won't cut it.

7. roryclear ◴[24 Aug 25 14:41 UTC] No.45004615{4}[source]▶
Maybe my view of frigate and tensorflow (assuming frigate still uses it) is outdated then. I’m referring to tinygrad vs tensorflow when I say GPU support, of course google’s tensorflow is best for google’s TPUs. I’ve had better luck using tinygrad on my personal devices, but I am biased as it’s been a while since I’ve used tensorflow
replies(1): >>45007150 #
8. sugarpimpdorsey ◴[24 Aug 25 15:54 UTC] No.45005215[source]▶
it's AGPL so you have to give anyone that views your camera feeds a copy of the source
replies(2): >>45005317 #>>45010617 #
9. ◴[24 Aug 25 16:06 UTC] No.45005317[source]▶
10. baruz ◴[24 Aug 25 17:07 UTC] No.45005848[source]▶
The author states elsewhere that the payments are for the use of their server, which can be reconfigured.
replies(1): >>45005897 #
11. diggan ◴[24 Aug 25 17:15 UTC] No.45005897{3}[source]▶
Yeah, I'll admit to not having tried Clearcam myself, I was just going by the information from the README as-is.
12. threecheese ◴[24 Aug 25 19:49 UTC] No.45007150{5}[source]▶
This would be a good point of differentiation to make on your GitHub page or for a technical audience on your website. Frigate is SOTA in many folks minds, and to show that you are using tinygrad over tensorflow may be a good “modern-ness” signal for that audience.

Edit: another solution in this space shows a list of supported ML runtimes, which would be good info for folks wanting to run on specific hardware. https://github.com/boquila/boquilahub

replies(1): >>45008521 #
13. roryclear ◴[24 Aug 25 22:52 UTC] No.45008521{6}[source]▶
Supported runtimes list would be nice, but I don't have access to much hardware to test on. I aim to remove most dependencies and support anything that can run tinygrad + ffmpeg
14. pimlottc ◴[25 Aug 25 03:06 UTC] No.45009819[source]▶
Sorry, which one are you talking about, frigate or clear cam?
15. skinkestek ◴[25 Aug 25 05:48 UTC] No.45010617[source]▶
Unless you’re making changes, isn’t it enough to just link back to the original repo?

That said, I’ve also been in the camp that avoids AGPL-except maybe as a way to sell a commercial license while still being "open source," or just to be obnoxious. And honestly, I am still failing to see the upside in being obnoxious for its own sake.

replies(1): >>45014651 #
16. roryclear ◴[25 Aug 25 15:03 UTC] No.45014651{3}[source]▶
It's AGPL because Ultralytics requires it to use YOLO: https://github.com/blakeblackshear/frigate/pull/10717

I'd make it MIT tomorrow if you know a workaround or alternative model

replies(1): >>45036814 #
17. skinkestek ◴[27 Aug 25 08:14 UTC] No.45036814{4}[source]▶
Thanks!

Both for the explanation and for being a person that respects the license even if you (like me) don't particularly like it.