5 Commits

Author	SHA1	Message	Date
raystlin	adc4dc0f74	Added new parameters to model input. ... New parameters have been added to define the input of the models: * ResizeOperation: by default center-crop was being performed, now it is configurable. * InputOrder: by default RGB was being used as the order for the array values of the input tensor, now it can be configured. * InputInterval has been changed to InputIntervals (an slice). This means that every channel can have its own interval conversion. * InputInterval can define now stddev and mean, because sometimes instead of adjusting the interval, the stddev and mean of the training data should be use.	2025-07-15 13:31:31 +00:00
raystlin	ad94d03576	Fixed classify and nsfw tests ... They were broken when the constructors changed to include metadata.	2025-04-13 14:53:34 +00:00
Michael Mayer	caf3ae1ab5	AI: Refactor use of face embeddings, labels, and nsfw models #127 #1090 ... Signed-off-by: Michael Mayer <michael@photoprism.app>	2025-04-10 16:49:34 +02:00
Michael Mayer	bfdb839d01	AI: Add TensorFlow utility package and improve model loading #127 #1090 ... Signed-off-by: Michael Mayer <michael@photoprism.app>	2025-04-07 05:26:45 +02:00
Michael Mayer	35e9294d87	AI: Add vision package and vision API endpoints #127 #1090 ... These changes allow to configure the computer vision models through an optional vision.yml configuration file. Note that the API endpoints are not yet functional and require further work. Signed-off-by: Michael Mayer <michael@photoprism.app>	2025-04-06 23:39:37 +02:00