If you plan on using the GPU for AI motion detection, then look into getting a Google Coral Tensor Processing Unit (TPU) instead. It is inexpensive, is better than a high end GPU at AI motion detection, and uses 2 to 4 watts of total power doing so. If this is accurate (no reason to doubt), it woul...

I see there's a lot of talk about leveraging a Coral TPU instead of a GPU for AI. I'm currently using a GTX 1650 for CodeProject AI and it's working very well, but would really like to to lower power consumption without sacrificing too much on AI detection. Which Coral TPU are people buying and whe...

The type of Coral TPU will vary based on your computer hardware. The two main types are the USB version and the PCIe version. You may need a NVme M.2 slot for the PCIe version. Make sure your M.2 slot is NVMe and not SATA. They can look the same, but I believe the current Coral TPU's require the NVM...

If you plan on using the GPU for AI motion detection, then look into getting a Google Coral Tensor Processing Unit (TPU) instead. It is inexpensive, is better than a high end GPU at AI motion detection, and uses 2 to 4 watts of total power doing so.

Adding a "me too". Android app has been very slow to load the last couple of months. I thought it was completely hanging, but it responds, it just takes 30s or so after opening before you can interact with it. Same here, but only when not on my LAN Wifi. Outside of my LAN it takes about a...

If you are going to use port forwarding it’s best to set this up your router itself, BI is probably using UPNP to set up forwarding and I assume that these are somehow being removed? Personally I use Tailscale for remote ipv4 access and ipv6 pass through where my mobile provider provides ipv6. I al...

I'm not sure if this is the same issue I am experiencing. I am using the Android Blue Iris app and noticed a change a month or so back. I can access my alerts and cameras fine at home (on my LAN) from the mobile Blue Iris app, but when I connect outside of home (cell or wifi), there is a very long d...

I'm quite impressed how low the timing is to analyse; my object detections are anywhere in between 100-200ms using the medium model. My setup is a virtual machine with 4x CPU cores Ryzen 5800U with the TPU using CPI-E passthrough in hyper-v, so in theory it should be pretty good and low overhead. I...

I'm quite impressed how low the timing is to analyse; my object detections are anywhere in between 100-200ms using the medium model. My setup is a virtual machine with 4x CPU cores Ryzen 5800U with the TPU using CPI-E passthrough in hyper-v, so in theory it should be pretty good and low overhead. I...

MikeLud wrote: ↑Thu Dec 14, 2023 2:47 pm The custom models are faster and have better accuracy. I use ipcam-general on all my cameras.
I am working on a Coral module that will work with custom models.

That would be amazing, good luck.