The Vision (Capability) Of The Cloud
Let’s discuss a few business models today!
Recently I’ve been fascinated by background removing services like:
PS: This does not mean I do not buy premium software, I’ve made a lot of stupid decisions on ones that offer lifetime subscriptions. It’s another high altogether to own something :)
Then there are features like Face recognition in pics that a lot of platforms like iOS, Android, Alexa, etc have been offering for a long time.
Let’s look at what the biggest players in the cloud have for us?
My objective is not to present a comparative analysis but to provide you with the features available in such a service. There would certainly be some differences between what AWS has to offer vs Google Cloud, I will provide some links if you wish to explore.
Both of these services offer Computer Vision capabilities to get information from Video & Images
Computer Vision is scientific feild that deals with how comupter can gain high level understanding from an image
What are the features that AWS Rekognition & Google Cloud Vision support?
- Content Moderation
- Face Compare & Search
- Face Detection & Analysis
- Lable common objects
- Customize Your Labels
- Detects Text From The Image
- Celebrity Recognition
- Video Detection
- PPE Detection
Understanding the uses of technology lets us come up with business models/ideas or problems that could be solved (these may not be out of the world ideas but its good to do some memory jogging and try to put 2 & 2 together) :
- In one of my past projects on a streaming (OTT) product, we wanted to have the UGC. The problem with UGC is monitoring the kind of content that could get uploaded. While working on the business plan, management came up with huge manpower & process expenses plus the time it would take to approve. The project was later called off BUT AWS Rekognition or Google Cloud Vision could have been a great module to implement. According to AWS companies like SmugMug, CBS & Mobisocial are using this effectively.
- I recently moved from iPhone 8 plus to 13 Pro, face recognition technology is old but I was new to it. While I liked it and started romancing it, I faced an issue when I went out. I was wearing a mask. With my old phone, the finger impression would have unlocked my phone so now I have to input my password. Since Rekognition & Cloud Vision does face detection & PPE detection, can these two features be combined? This will help us unlock our phones even if we have our masks on.
- How about a fun app to compare faces? There is always a debate about which parent the child looks more like that could be solved! People always called my younger sibling and me twins, well with this technology we could now tell everyone how similar or different we are. There are apps in this space like, Face/Face, Face Similarity, Face Comparision Slider, Gradient, Star by Face, etc
- Have you noticed ads on OTT platforms especially YouTube on longer videos? If there is a long video with multiple songs in it, somehow YT already knows when a song ends and the new one begins and inserts an ad in-between without disturbing the song. I used to wonder how do they figure it out! Now we know :) The technology works on process build API with Machine Learning. Each video detection could return SMPTE (Society of Motion Picture and Television Engineers) timecodes and timestamps i.e now we know the exact frame of the video we want to play around.
- We can use object detection effectively as well — services like Google Lens, CamFind, Amazon Shopping, Pinterest, ASOS, etc use object detection within an image and use it for eCommerce. Eg: point it to a shirt in a picture & will give you links to where you can buy it from.
We can go on & on. Ideas and implementations are endless, they are, as always, just limited by our imagination.
PS: As promised earlier Cloud Academy provides are a good comparison between AWS Rekognition & Google Vision
If you have more ideas feel free to share them.