AWS Panorama - Part II, unboxing, review and guided setup.

[@Datasmum]

So great Mike! Also so good at the end when a car drove down the street and it detected it!

[@mikegchambers]

Thanks! That was a nice little addition wasn’t it. :)

[@selimcansakar8251]

Great video ser! Ps: Instead of ssh tunneling and paying for the EC2, you can also build them on Ubuntu 20.04 with Parallels on M1 chips since all you need is ARM64 🙃

[@mikegchambers]

Nice, I’ve not tried that and I don’t currently have an M1. Thanks for the tip.

[@shaoyongshen5238]

excellent! thanks to Mike

[@sikretp]

Rewarded 😍

[@chetanmundhe7899]

It's awsome...very usefull. Just would like to see if we can generate customer analytics report like their sentiments analysis report or emotion detection, pass through rate for different section in area. Can you extend these video series further to get chance to create the analytics dashboard. It will be very usefull

[@dishitamehtalia5385]

Hello Mike, In this people counter application, what additional changes I should make if I want to add age and gender feature for the people detected? Thank you in advance

[@mikegchambers]

Interesting question. You would need a model or more likely a couple of different models that make those detections. The Panorama itself can host more than one model. You might find commercial datasets/models that could make those predictions.

[@dishitamehtalia5385]

@@mikegchambers Thank you for your instant reply. I have actually found a caffe model for age and gender detection but I am unsure about how to add those model to current people counter application since I am new to Panorama environment. Is there any document which could help me understand on how to add those models?

[@mikegchambers]

@@dishitamehtalia5385 it’s gonna take some fiddling about and you might have to compile the model for the device. The best place to look is there doco here: github.com/aws-samples/aws-panorama-samples and dig in to the examples. But… maybe this is where you’re already at. Basically once you have come to grips with the Panorama and it’s SDK, the rest is down to some ML Python code to glue it together.