Open goern opened 4 years ago
@goern LGTM For an end to end computer Vision demo we need to define the data streaming architecture. The current architecture is focusing the MQTT and sensor data. Do you have a recommendation for TF based predictions i.e. form Kafka / Kafka Streams?
@pacospace dont we have something like that in the AICoE? By the AIOps folks?!
@stefan-bergstein for a computer vision related demo, cant we simply take something like https://cdn.pixabay.com/photo/2013/09/28/11/34/tablets-187704_960_720.jpg copy it 500 times and use it as a training set? create 5 anomalies and go ahead which this kind of a dataset?
Cc: @pacospace