Does Azure Anomaly Detection API works for Vision data

Question

I have tried Azure Anomaly detection API with Sample dataset provided in the Github.

I'm able to setup the Python environment and run with the sample data provided.

Now, I'm planning to use the Anomaly services for detecting anomalies in Videos., Does azure anomaly services support vision time series dataset.

Answer 1

You can also use built-in A/D in Stream Analytics if you are putting together a hotpath/real-time analytics pipeline with PaaS services stitched together consisting of Event Hubs or IoT Hub as the ingress sources and Stream Analytics as processing engine feeling any other relevant services downstream.

Here are other that can help.

1. Time series anomaly detection using LSTM autoencoder sample using Keras created: https://github.com/johnwu0604/time-series-anomaly-detection/blob/master/Time_Series_Anomaly_Detection.ipynb

2. AutoML based sample (somewhat outdated, will need some refresh): https://github.com/deepamu/automatedml/blob/master/Predictive%20maintenance%20NASA%20sample.ipynb. Modelled as a Regression problem but you could use time-series forecasting for same scenario as well.

3. Near real time Anomaly Detection using Cognitive Services: https://github.com/jomit/anomaly-detection