Snapchat, the popular Social app, launched SnapML last June: an important update to its development tool (Lens Studio) that allows the use of Machine Learning algorithms to create Lens, that is filters that enrich the user experience .

Snapchat is not new to this type of initiative having always been a pioneer in the adoption of innovative tools or new technologies, just remember for example the Spectacles: glasses on which to use Augmented Reality applications (2016).

Snapchat Spectacles
Snapchat Spectacles
Photo credits — Wareable.com

More recently, October 2020, they have included in Lens Studio, full support for the iPhone 12 Pro Max LIDAR (which I hope to cover…


From photos to rendering — Credits NERF-W Project

Here I will tell you about “NERF in the wild” a research presented in August 2020 and which has all the features to revolutionize some application areas starting from ** Augmented and Virtual Reality **.

The aim of the research is to produce a 3D visual synthesis of places starting from photographs of the same place very different from each other, taken at different times and with the automatic removal of objects / people that are not relevant to the object (done by the neural network itself during the process).

Eg. the video of the Trevi fountain below was generated…


(Photo Credit — Karras+ Arxiv 2019)

GANs are neural networks capable of automatically generating content (audio, video, photos, etc.) from starting data consistent with those that are intended to be generated.

In practice, they can be used, for example, to create photographs of totally invented subjects (humans, animals, etc.), to artificially increase the data to be used during the training of other neural networks (with benefits on effectiveness, bias reduction, …).

But you can also think of different application scenarios for example related to advertising, posts on Social Networks, cinema, etc. 🎬

Their architecture is very interesting: there are two networks, a Generator and a Discriminator……


(Photo Credit — Karras+ Arxiv 2019)

Le GAN sono reti neurali in grado di generare automaticamente del contenuto (audio, video, foto, etc) a partire da dati di partenza coerenti rispetto quelli che gli si vuol far generare.

In concreto si può usare, ad es., per creare fotografie di soggetti totalmente inventati (esseri umani, animali etc), per aumentare artificialmente i dati da utilizzare durante il training di altre reti neurali (con benefici su efficacia, riduzione bias, etc).

Ma si possono pensare anche a scenari applicativi diversi ad esempio legati a pubblicità, post su Social Networks, cinema, etc 🎬

La loro architettura è molto interessante: esistono due reti…


Empowered Retail @ Milano Digital Week 2019

For the second year in a row Milano city organised the week of the digital (aka Milano Digital Week): a 7 days of events all across the city with the goal to share know-how, experiences and projects related to Artificial Intelligence, Big Data, Robots, Blockchain, etc. …


Just Commerce è da qualche giorno disponibile sul mercato. Il nome sintetizza a nostro parere perfettamente la nostra visione: non ha più senso parlare di e-commerce e vendita da negozio fisico, gli strumenti, la cultura, etc ci stanno sempre più portando ad utilizzare in maniera olistica tutti i canali possibili durante il processo di acquisto (aka customer journey). E’ quindi naturale conseguenza prendere atto che questo approccio omnicanale, in cui il negozio fisico ha l’opportunità di ricoprire un ruolo determinante, sia l’unico modo con cui si acquistano beni/servizi: it’s just commerce, per l’appunto.

Frutto dell’esperienza pluriennale del nostro team, è…


Ultimamente, lavorando nel cercare di far guadagnare un nuovo ruolo ai negozi nell’ambito di questo scenario estremamente competitivo che si è venuto a presentare, più volte mi è stata posta la domanda:

“come possiamo competere con Amazon?”

Dove al posto di Amazon ci si può mettere un qualunque altro player digitale analogo (dove, ok, il termine analogo non è di facile applicabilità ad Amazon :-)).

La risposta è nell’articolo che segue, che sintetizza il nostro pensiero in Stentle.

Se vi interessa capire come meglio si possa applicare nel vostro caso scrivetemi pure su alexio.cassani@stentle.com che sarò felice di parlarne.


Nothing really authentic to say, but I like to share with you this article that sounds pretty interesting if you have some Java/Android skills. Apparently it’s possible to reverse the APK and obtain a close-to working source package of PokemonGo for Android:

A couple of interesting links related to the same track:

and the official list on Reddit of the group that is trying to build an API to gather information about players (sounds creepy IMHO):

If you have more to add please be free to comment this story.

I think I’m done.

Alexio Cassani

Tech Entrepreneur, Co-Founder and CEO of Stentle.com (a M-Cube Group company since 2019) — AI Advisor - Retail Transformation & E-Commerce Expert

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store