FARBE

Technologies

Combined Experience in

  • 40+YearsIndustry
    R&D
  • 50+YearsAcademic
    R&D
  • 15+YearsProduct
    Development

Global Reach

US | LK

Contact Us
Tell Us What You Need

Core Technologies

Engineered for the Real World

Farbe’s technologies are engineered to solve practical problems across industry, healthcare, energy and agriculture. From spectral imaging for quality and food safety assessment, to wearable sensing for sports and rehabilitation, to computer vision for inspection and monitoring, and intelligent analytics for energy systems, each capability is built for deployment in real operating environments. Backed by expert scientific leadership, every solution can be tailored to your requirements, and you can own and commercialize the result.

Spectral Imaging

Our in-house multispectral imaging system captures information that standard cameras cannot see. Supporting both reflectance and transmittance imaging, this low-cost, modular device offers a practical alternative to commercial systems, enabling rapid, non-destructive, data-driven quality assessment for agriculture, food safety and material analysis in field and laboratory settings.

  • Soil texture estimation (USDA classes)
  • Aflatoxin contamination detection in food
  • Reheated coconut oil identification
  • Detecting adulterants in turmeric
  • Hyperspectral unmixing with spectral variability

Smart Energy

Our smart energy research turns everyday infrastructure into intelligent, self-sustaining systems. From a single meter, lightweight AI breaks consumption down to individual appliances, while battery-free indoor IoT nodes harvest light and share power through optical wireless transfer. Digital twins and short-term solar forecasting help plan generation, demand and crop growth across homes, industry and agriculture.

  • Appliance-level monitoring from one meter (NILM)
  • Lightweight AI for smart meters
  • Battery-free light-powered indoor IoT nodes
  • Digital twins for tea growth
  • Optical wireless power transfer between nodes
  • Short-term solar irradiance forecasting

Wearable Motion Capture

We have built a wireless, ergonomic, low-cost wearable platform using inertial sensors to capture human movement beyond the laboratory. It enables accurate joint-angle estimation and movement analysis for healthcare, rehabilitation, footwear research and sports performance, with a general-purpose design ready to extend to high-speed sports like athletics and cricket.

  • Joint-angle estimation via sensor fusion
  • Gait and posture analysis
  • Inter-joint coordination and strength insights
  • Footwear and terrain effect studies
  • Objective athlete performance benchmarking
  • IMU and video weightlifting analysis

Computer Vision

We build lightweight, efficient and interpretable vision systems for detection, tracking and scene understanding. Designed for low-power edge hardware, our work includes group activity recognition, energy-aware multi-camera sensing and image-based solar forecasting, supporting real-time, battery-powered applications in robotics, drones, smart surveillance and plant monitoring with continuous environmental awareness.

  • Multi-person group activity recognition
  • Low-power wide-view motion monitoring
  • Foveated, energy-aware multi-camera vision
  • Solar forecasting from all-sky images
  • Fisheye distortion correction for clouds
  • Image-based plant growth phenotyping

Generative AI

We advance generative AI through new algorithms and careful study of real-world behaviour. Our research spans diffusion models, implicit neural representations and large language models, with applications in high-fidelity image generation, heritage restoration, medical imaging, computational biology and responsible, trustworthy AI, including bias, safety and hallucination detection.

  • High-fidelity diffusion-based image generation
  • Medical image augmentation, gene-expression imputation
  • COSMO-INR overcoming spectral bias
  • Historical image colorization and restoration
  • Detecting LLM hallucinations and misinformation
  • LLM bias and fairness evaluation