Frigate+ questions #11988
-
What is your question:I was reading the Frigate+ docs and searched the forum here but some things are not perfectly clear to me. I'm trying to understand:
Thanks in advance. |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 3 replies
-
The models only support a specific defined set of objects. You can't add custom objects nor does it support all of the labels available in the default model. Here is a quote from the docs: "Frigate+ models support a more relevant set of objects for security cameras. Currently, only the following objects are supported: person, face, car, license_plate, amazon, ups, fedex, package, dog, cat, deer. Other object types available in the default Frigate model are not available. Additional object types will be added in future releases."
This is because many users have cameras that look different enough that the base model won't perform well without some tuning. I don't want people to think they can just use the base model and get good results. Another quote from the docs: "The baseline model isn't directly available after subscribing. This may change in the future, but for now you will need to submit a model request with the minimum number of images."
I picked a limited set of objects to start with and will be adding more in the future. I didn't want to start out by having users label objects that wouldn't be supported in the models they requested.
I don't understand the question. Every model you request in Frigate+ will support all of the labels listed above. From the docs: "The images you upload are used to fine tune a baseline model trained from images uploaded by all Frigate+ users. This fine tuning process results in a model that is optimized for accuracy in your specific conditions." and "This model will support all label types available even if you do not submit any examples for those labels." |
Beta Was this translation helpful? Give feedback.
-
|
Thanks a lot for your clarifications. Do you plan to allow users to add custom objects anytime in the future? |
Beta Was this translation helpful? Give feedback.
-
|
I would like to. It will require a lot of changes, so for now I am focused on making it easier to expand a common set of labels. Custom objects will also require a lot more examples. You might need to provide hundreds or thousands of examples to get good results for something since you don't the collective benefit of pooling examples across users. |
Beta Was this translation helpful? Give feedback.
-
|
I have another question about Frigate+ (please let me know if I should open a new discussion for this, or if it has already been answered): is there a limit on the number of cameras, or the images I upload and label per camera, when creating a model? |
Beta Was this translation helpful? Give feedback.
-
|
No, there are no camera or image limits (as a max). The only limits are that you must upload at least 10 images between each model |
Beta Was this translation helpful? Give feedback.
The models only support a specific defined set of objects. You can't add custom objects nor does it support all of the labels available in the default model. Here is a quote from the docs:
"Frigate+ models support a more relevant set of objects for security cameras. Currently, only the following objects are supported: person, face, car, license_plate, amazon, ups, fedex, package, dog, cat, deer. Other object types available in the default Frigate model are not available. Additional object types will be added in future releases."