In this article we are going to peek behind the fashion industry and see what brands trends and style, so we could what AI can improve.
Hi, Happy new year !!!
Before we start, take a moment to appreciate another year that was given to us. Now we can dive in our first 2019 article.
AI has been revolutionizing one industry at a time and we have seen a significant rise in efficiency, quality and supply. Last year alone many companies emerged and some pivoted to include AI in their products. Some examples can be:
- Cocoon Cam Clarity baby monitor — This smart baby monitor camera can tell if the infant is breathing and send users real-time data and alerts. It also allows a two-way interaction with the baby without the need for any wearable or garment attachments. Machine learning algorithms and HD camera mounted above the baby’s cradle make this possible.
- Deepmind AlphaFold — A system that generates 3D models of proteins that are far more accurate than any that have come before — making significant progress on one of the core challenges in biology.
This is just the tip of the iceberg. Deep learning a subset of Machine learning is the major contributor for most of these breakthroughs made possible in 2018 because it can learn complex representation from large data. We can clearly see the AI can change a variety of industries such biomedical, agriculture to robotics and many others. World domination!!!🤣🤣🤣
Now with all that said, what can AI do for the fashion industry in 2019.
First, let’s understand what is the heart of the fashion industry — this industry is well known for displaying and commercialize luxury, power and extravagance. Who doesn’t like to stand out?
There some principles and techniques that companies use to not only stand out but to differentiate themselves from one another, that's where branding is born. Branding begins and ends with a logo, fashion brands communicate their uniqueness through a wide range of visual cues such as colour, patterns and shapes.
What if we could teach a computer to learn and personalize user experience?
There has been a shift in the fashion industry and a lot of research and implementation are being done with big brands like Nike, Adidas, Burberry, and Levi’s that are implementing AI in everything from personalized running shoes to chatbots helping the shoppers.
What about enabling computers to understand this visual queues?
Computer vision, in this case, is our go-to field. The image above is a close enough and vivid representation of how computers can see and discern the differences between brands.
If you are a geek you can access the notebook and play with it here.
Amazing to see the world from computers point view.
With the most recent success of computer vision and the rise of online commerce, there is a huge excitement to turn computers into visual experts. The everchanging landscape of the fashion industry has provided a unique opportunity to leverage computational algorithms on large data currently available to achieve knowledge and expertise unattainable for any individual fashion expert.
According to this paper, a study was conducted asking 5 people on Amazon Mechanical Turk to label each product image in the bags category according to the visibility of the logo into one of three groups: (i) Logo (ii) Repeated Logo when a pattern of repeated logos or monogram and (iii)No logo. 46% of the dataset images contained a logo and 51% contain no logo. The classifier they used correctly predicts 65.01%, 68.67% and 54.46% of the brands in (i), (ii) and (iii) respectively. This is very significant, given that group(iii) constitutes the majority of the dataset and confirms that deep classifiers learn unique visual characteristic of all three groups.
There are some quite interesting previous researches that tackled outfit compatibility and recommendation, style and trend recognition.
Leveraging computer vision we can quantify deep representations to analyze and interpret visual characteristics of fashion brands. Literally knowing what techniques fashion companies use to make their brands stand-out, we also can know the correlation of parts of the algorithm that detect logos images of fashion items and the parts of the system that detect patterns.
Key points to remember:
- AI is changing many industries
- The fashion industry can use AI to create better products and customer experience
- AI can also surpass a human expert at detecting brands even when there is no logo, through patterns recognition.
For you that made it till the end here is a bonus custom made picture:
THANK YOU FOR READING!!!
You are amazing if you enjoyed please leave as many claps as you can because I do this for you — this year I have amazing projects and content lined up, follow me if you don’t want to miss my new articles every Wednesday.