toz

my new compreface instance is ver 0.61 (jumped from 0.6.0). not sure if its from the new code, but if it try to train my images with this new gpu-model...

weird. on my first seperate CPU maschine i got now cuda-errors. but on this maschine is no nvidia-gpu. how did this switched over? ``` compreface-core | 2021-11-07 18:22:22.181578: E tensorflow/stream_executor/cuda/cuda_driver.cc:351]...

still interesting modell-zoo for exploration of more accuracy and maybe faster detection: https://github.com/deepinsight/insightface/tree/master/python-package#model-zoo

thanks for pointing out. i will observe the mqtt if i find clear critera to catch misses. i must start with a positve-match == the misses is empty. 

i forgot about the api-service-url. GREAT. new ideas coming up... 👍 thanks a lot for bringing this up ... i will test the originating problem a bit more to find...

looks like mqtt sends only matches... got misses on the gui, but had nothing on the mqtt-topic... may you could check if sending is filtered only for matches?

i still want to catch all misses and unknown together (all faces marked RED in the gui). is there an easier way than filtering for borth "misses" && "unknown" array?...

there are mqtt-messages missing, while i am experimenting to find the right node-red flow. i see the gui filling up, but not the same mqtt-messages with unknown/missing .. can this...

All the RED-ones are missing in mqtt, but thats the ones i want to catch. i assumed they are also send through mqtt, even when their box-area is too small....

may try a different model with a custom compreface docker? -> https://github.com/jakowenko/double-take/issues/151