Vision Intelligence: TYQ-i

You are Industry 4.0 ready. So are we.
TYQ-i - Vision Intelligence for the Smart Factory.

Every factory floor is looking to improve efficiencies - be it reducing wastage, controlling yield, optimizing expensive tool usage or quite simply improving finished product quality. TYQ-i is Ignitarium's Visual Deep Learning based defect detection platform that can vastly improve your Quality Assurance process.
Keep your factory line humming with TYQ-i.

Key Features of TYQ-i

  • Vision Intelligence based industrial defect detection
  • Hybrid implementation leveraging classical CV and ML techniques
  • In-field Trainable
  • Custom Neural Network that minimizes the requirement for training data sets
  • Solution realizable on cost-effective embedded platforms
Vision Intelligence for the Smart Factory

TYQ-i Product

Industrial defect detection and quality assurance


TYQ-i is a distillation of Ignitarium's deep experience in AI-for-vision. At its core, it blends the best of classical CV techniques with sparse custom neural nets to create an efficient hybrid solution for mission-critical industrial use cases.

Visual Deep Learning - Minimal training images

Minimal training images

Advanced custom (and not pre-trained open) CNNs that can be trained for high accuracy inference with minimal data sets (often less than a few 10s)

Visual Deep Learning - Scalable across defect classes

Scalable across defect classes

Common Neural Network platform architecture that can detect edge, surface, color or dimensional defects

Visual Deep Learning - Integrated SDK-based solution

Integrated SDK-based solution

Efficient SDK to allow quick deployment of the solution in live environments, irrespective of defect class being targetted

Visual Deep Learning - Works with low-cost cameras

Works with low-cost cameras

Does not require high-end purpose-built industrial grade Machine Vision cameras. Can even work with regular webcams.

Visual Deep Learning - In-built benchmarking famework

In-built benchmarking famework

The TYQ-i framework has a native benchmarking framework that allows quick identification of bottlenecks in the pipeline - be it registration, classification, inference or sub-stages in between.

Visual Deep Learning - FPGA-acceleratable architecture

FPGA-acceleratable architecture

Efficient architecture that can seamlessly transfer critical functions from the Vision processing pipeline into low-cost FPGAs to achieve higher frame rates or reduce latencies.

Use cases

TYQ-i - Edge & Surface Defect Detection

Edge & Surface Defect Detection

Detection of Edge and Surface deformities in Metal washers traveling on a conveyer belt

Read More
TYQ-i - Missing Artifact Detection

Missing Artifact Detection

Detection of presence or absence of Nuts and Bolts on designated slots of machine parts.

Read More
TYQ-i - Real Time Color Detection

Real Time Color Detection

Color detection - including shade variations - in real time.

Read More

TYQ-i Demo

Edge & Surface : Metal Washers

TYQ-i in action : Edge & Surface Deformity Detection in Metal Washers

Visual Deep Learning - Edge & Surface : Metal Washers
Visual Deep Learning - Edge & Surface : Razor Blades
Visual Deep Learning - Missing Artifacts : Nuts & Bolts
Visual Deep Learning - Color Detection