Mindtech Raises $3.25M to Train CCTV with Synthetic Humans

A Future of Privacy: Synthetic Data for AI Training

Envision a future where personal privacy remains inviolate, facial recognition databases cease to expand, and privacy regulations are consistently upheld. This reality is drawing nearer. Could businesses transition away from conventional CCTV systems and instead utilize synthetic humans to simulate countless potential scenarios?

This possibility is now being explored by Mindtech Global, a U.K.-based startup that has recently secured substantial investment. The company has created an end-to-end platform for generating synthetic data.

The Chameleon Platform

Essentially, Mindtech’s system can generate realistic visual scenarios, such as pedestrian behavior at a crosswalk or customer actions within a retail environment. This generated data is then employed to train AI-powered computer vision systems for a diverse clientele, including retailers, warehouse facilities, healthcare providers, transportation networks, and robotics companies.

The company recently completed a $3.25 million early-stage funding round, spearheaded by NPIF – Mercia Equity Finance, with additional contributions from Deeptech Labs and In-Q-Tel.

The involvement of In-Q-Tel is particularly noteworthy. This organization invests in startups that bolster U.S. intelligence capabilities and is headquartered in Arlington, Virginia.

Mindtech’s Chameleon platform is engineered to enhance a computer’s ability to interpret and anticipate human interactions. Traditional AI vision system training relies on real-world data, often sourced from CCTV footage.

Addressing the Challenges of AI Training

This process is often hampered by privacy concerns, and can be both expensive and time-intensive. Mindtech asserts that Chameleon resolves these issues, allowing customers to rapidly “create unlimited scenes and scenarios utilizing photo-realistic, intelligent 3D models”.

Furthermore, the use of synthetic humans enables the training of AI vision systems to mitigate biases related to diversity and inclusivity.

According to Steve Harris, CEO of Mindtech, machine learning teams can dedicate as much as 80% of their time to data sourcing, cleaning, and organization.

“Our Chameleon platform addresses the AI training challenge, allowing the industry to concentrate on more valuable tasks, such as AI network innovation,” Harris stated. “This funding will facilitate our growth and enable the development of a new generation of AI solutions that possess a deeper understanding of human interaction with the world.”

Real-World Applications

Consider these examples: A child momentarily slips away from a parent in a shopping center. The synthetic CCTV, trained extensively within Mindtech’s simulated environment, can identify this situation in real-time and alert staff.

Another scenario involves a delivery robot encountering children playing in the street, enabling it to navigate the situation safely.

Finally, a passenger exhibiting erratic behavior near train tracks can be automatically detected by the CCTV system, triggering an immediate response.

Nat Puffer, Managing Director (London) at In-Q-Tel, commented: “Mindtech’s Chameleon platform demonstrated a high level of maturity and existing commercial success. We are enthusiastic about the platform’s diverse applications and its potential to overcome a significant obstacle in the creation of more intelligent AI systems.”

Miles Kirby, CEO of Deeptech Labs, added: “We support ambitious teams with groundbreaking solutions. Mindtech’s experienced team is dedicated to revolutionizing AI training, and we are pleased to support their progress.”

Potential and Considerations

While the technology offers numerous benefits, potential applications could include identifying shoplifting or optimizing warehouse worker performance. However, Mindtech’s platform theoretically allows customers to eliminate managerial biases and improve customer service.