Synthetic media is out of the research papers and going mainstream. In the last year, a handful of companies have made the technology accessible via innovative applications. Bria, an Israel-based start-up, is one of the newest addition in this field, recently launching with revolutionary AI-assisted creative tools. But it is just a start. We sat down with Co-founder and CTO Yair Adato to learn more:

 

 – A little about you, what is your background?

portrait of Yair Adato Co-Founder and CTO at Bria
Yair Adato, Co-Founder and CTO at Bria

My passion is bridging new, state-of-the-art technologies and the world of business. To see how we can close gaps where technology is not yet fully mature, and businesses don’t understand exactly how, where, and when to use it.

I’ve been in the field of computer vision for years. I’m fascinated by all aspects of visuals – how the human brain and computers make sense of them. It was the focus of my Ph.D., which consisted of joint research between Ben Gurion University in Israel and Harvard. I continued with this work for a while at Google. 

Nine years ago, I joined Trax, which at the time was a small start-up with around 15 employees. I helped grow the company into a unicorn with a staff of over 900, serving as its CTO and VP of R&D. 

When academic research began to talk about a new concept of generative AI, I was inspired by how this could transform the world of creativity. This led me to found Bria two years ago.

Outside of Bria, I live in a small village in central Israel with my wife, two children, a cat, and a dog.

Screenshot of the The Bria web interface.
The Bria web interface.

– What is Bria? What does it solve?

Bria’s vision is to enable anyone to instantly create any visual, still or video, at scale – and without the need for a camera or Photoshop. From thoughts to visuals in a few seconds. We believe that this is the next major jump in the evolution of visual communication. 

There’s so much creative potential in the world that’s untapped. Because to create a visual, you need technical skills and/or a budget. You need to be able to draw, design, or use complex software like Photoshop or, alternatively, to have long, slow, and expensive production days. 

Bria empowers anyone to unleash their creative potential. You tell Bria what visual you’re looking for, and it generates various photographs and videos which match your request. You can then modify and customize them if you want to make any adjustments. And you can do all this without any technical knowledge.

We believe that in the near future, the best creators will be the people with the best imaginations – not the best technical skills or the biggest budget.

– Is it an app, a SaaS or an API ? How best can users interact with it?

Bria is a B2B or B2B2C SaaS company. With the broad vision to revolutionize visual communication, we want users to engage with our platform directly, and we want to enable other companies to integrate it into their products and activities.

The Bria API is for enterprises who want to integrate the creative power of Bria into their own platforms. They can offer it to their users or use it internally to improve their visual performance. We see a lot of potential here with image banks, video platforms, photo editing tools, content platforms, eCommerce platforms, marketing automation, and more.

Bria’s online platform is a web app that’s currently open to anyone to use for free – we want to give everyone the power to unleash their creativity. 

– Photo editing tools is a crowded space with many start-ups involved. How does Bria create room to live and grow in?

Bria is a photo (and video) generation tool – not a photo editing tool. We replace the camera, not the editor. We don’t have pixel editing capabilities, layers, or paintbrushes. Moreover, we already have several photo editing tools as our clients. A good photo editing tool must have generative capabilities inside the tool, and this is where our API plays a crucial role.

We offer a unique solution that creates a photo-realistic visual precisely according to your needs from scratch, with no camera or Photoshop needed.

–  Who do you foresee being your primary users? People who create and sell images like Photo enthusiasts, professional photographers, or people who use images like pro image buyers and marketers? 

Visual communication is something we all do all the time. It’s something that everyone needs and wants to do to the best of their abilities. Anyone who has a need to create and modify visuals can use our platform without any special technical skills. 

This new creative independence will lead to an explosion in creativity. Removing time, skill, and resource-related barriers will give more people the opportunity to take more risks with their creatives. They’ll be able to do more with each visual – creating infinite variations, transforming images into video, and more. The only limit to your creativity will be your imagination. 

Specifically, we see a real pain point with marketers who have a constant need for more and more visuals – but time, talent, and budget constraints force them to compromise between getting the result they want and getting their communication to the market in a timely manner.

– How does Bria leverage synthetic media and make it indispensable?

Creating synthetic visuals that look completely real is a significant challenge. It’s even harder for visuals containing people due to the uncanny valley effect. 

This effect describes how people react to artificial likenesses of humans. In general, the closer the likeness is to a real human, the more positive our response to it. The most positive empathetic response is to a likeness that we can’t distinguish from reality.

But, just before that indistinguishable likeness, there’s a hugely negative emotional response to something that looks very close to being human but is just a bit off. This is called the uncanny valley.

A graph representing the uncanny valley illustrated with stills of cartoon movies
The Uncanny Valley Image as per BLOOWILLBOOKS

A great example of this is a CGI-animated movie called The Polar Express. There was a very negative reaction to it, with many people saying that the characters gave them the creeps. They’d fallen headlong into the uncanny valley. 

A polar express stills of 2 characters of the movie looking human but not quite
Too close for comfort?

After some small changes to the hair, eyebrows, nose, and lips, the boy looks a lot more familiar.

An editied version of the previous image where the character looks more comfortably real
THE POLAR EXPRESS film still courtesy of Ward Jenkins

So there is a very slim margin of error. We have to create visuals that are spot-on – or risk falling into the uncanny valley from which it is very hard to recover. We’ve focused a huge amount of effort on developing a patented generative process that’s designed to prevent the uncanny valley and deliver perfect results. 

In addition, we took our strong multidisciplinary team with researchers from computer vision, natural language processing, machine learning, and computer graphics – and we enriched it with experts in psychology who understand how we view and make decisions about visuals. Together, we’ve built a patented QA process to ensure that Bria’s synthetic visuals pass the ultimate test – the human eye. 

–  What are the moral/ ethical limitations of synthetic media offering? What is the line Bria (and no other) should cross?

I recently wrote that those of us who are leading this revolution need to lead the conversation about exactly this topic. At Bria, we have a VP of Ethics, and from day one, we’ve been building the product while facing up to these hard questions. 

Generally speaking, visuals can be used either to capture a memory or as a tool to tell a story, a piece of reality vs. visual communication. We’re focusing on the latter, and this is why we are not a B2C company. Consumers use visuals mostly for memories, while businesses use visuals mostly for marketing and narrative building. 

– The platform just launched at the beginning of March 2022. Can you share some insight on the next steps? What should we expect to see released from Bria in the near future?

By the end of 2022, Bria’s API will be integrated into several leading platforms. We have a really exciting pipeline that we’re working on, including video generation and brand book adoption. There’ll be a series of product releases over the coming months that will contribute to our vision of empowering creative people with creative independence. 

Bringing stills to life with dynamic video generation

 

Applying brand book guidelines to visuals at scale

 

– Finally, in your wildest dream, what would you like to see Bria offer that technology cannot yet deliver?

I have a dream, and it’s not so wild. I want anyone to be able to use Bria to create any visual from scratch and then modify the visuals to explore their creativity. The dream will become a reality before too long, and the way we all think about visuals will change forever.

Author: Paul Melcher

Paul Melcher is the founder of Kaptur and Managing Director of Melcher System, a consultancy for visual technology firms. He is an entrepreneur, advisor, and consultant with a rich background in visual tech, content licensing, business strategy, and technology with more than 20 years experience in developing world-renowned photo-based companies with already two successful exits.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.