Five Learnings From 15 Years In Perception

How has perception and computer vision changed in a decade and a half?

Adam Rodnitzky

What have I learned in fifteen years?

In the fall of 2008, I was working on my third startup, ReTel Technologies. Our goal was to analyze shopper behavior in grocery stores, and use that data to help stores and brands improve the customer experience and store profitability. But we had a challenge: how do you anonymously track hundreds of shoppers per day in a store? We thought we had the answer: active RFID tags on every shopping cart.

Since that lucky discovery of OpenCV in the late naughts, and having spent over fifteen years working in perception and computer vision, I feel qualified to share some observations about these transformative technologies and their place in the world. Let’s dive in!

Observation 1: Computer vision has been consumed by AI.

Today, that script appears to have flipped. While there are still many companies and research labs focused on core computer vision tasks, there are now orders of magnitude more that are focused on AI-driven applications.

Observation 2: Some of the most well-entrenched problems could be solved through AI.

Do we need more research into scene segmentation or facial feature tracking? Sure, but industry would likely benefit from more attention to challenges that have yet to be sufficiently attacked.

Observation 3: Some of the most well-funded robotics startups have only deployed a handful of devices.

Building a robot brings a completely different set of risks, timelines, and capital requirements. Robotics companies have traditionally required larger amounts of capital and time to get to market.

Observation 4: First-time deep-tech founders insist on building everything from scratch.

Some founders tend to aggressively avoid third-party tools, which can lead to inefficiencies. It’s essential to strike a balance for quicker, cost-effective building.

Observation 5: Perception and computer vision is now pervasive, and will only become more so in the future.

Fifteen years ago, software-addressable cameras were rare. Today, nearly every new car sold is equipped with multiple cameras and sensors, enhancing safety and the driver experience.

Conclusion

After a decade and a half in the perception and computer vision space, I only see it continuing to accelerate in its importance and richness of applications.