AI-powered image analysis tool for defect detection in industrial environments, with a guided journey for customizing new defect types.

A training pipeline for customizing YAMNet to recognize new environmental or industrial sounds.

On-device fine-tuning/validation of ResNet classifiers—customize without cloud dependency.

Guided training environment for creating and preparing custom Face Recognition models, with optimized deployment and inference on SECO hardware.

Energy-efficient object detection optimized for edge platforms with power and thermal constraints.

Bridging the Field and Digital: Universal Connectivity, Zero Latency, and On-Device Machine Learning for Industry 5.0.

Detects face presentation attacks and verifies liveness from RGB images to strengthen biometric authentication systems..

Real-time detection of critical alarm and emergency sounds for industrial safety applications.

Real-time semantic segmentation for understanding roads, vehicles, pedestrians, and other elements in visual scenes.

Detects faces and performs face recognition using a lightweight Edge Impulse pipeline designed for edge devices.

Classifies physical activities in real time from wearable motion sensor data.

Predicts ground settlement around tunnel boring operations using geotechnical and TBM operational data.

High-resolution neural network for accurate keypoint detection in human pose estimation tasks.

Recognizes hand gestures over time to enable touchless human–machine interaction on edge devices.

AI vision application to detect human falls in real time for safety, elderly care, and industrial risk monitoring.

Real-time image classification for recognizing visual categories on edge devices.

Detects and maps land-cover changes from multi-temporal satellite or aerial imagery for environmental monitoring and urban analysis.

Deep learning system to upscale and enhance low-resolution images while preserving fine details.

Efficient single-camera depth estimation using quantized MiDaS v2 model optimized for edge AI deployment.

Detects and locates objects in images using the efficient EfficientDet architecture optimized for edge deployment.

Detects and locates multiple object categories in images and video streams using the lightweight YOLOX architecture.

AI-assisted HMI that detects operator presence during machine alerts and provides contextual troubleshooting information.

Detects manufacturing defects in PCB images for automated visual quality inspection on edge devices.

Lightweight and quantized PoseNet model for real-time multi-person pose estimation on edge hardware.

Real-time people detection and counting for industrial and monitoring applications.

Detects and recognizes text in complex real-world images using a high-accuracy OCR pipeline for edge applications.

Identifies or verifies speakers from voice recordings using robust ECAPA-TDNN speaker embeddings.

Short-term electricity consumption forecasting based on historical time-series data.

Detects and recognizes text in images using a lightweight OCR pipeline optimized for edge devices.

Classifies human actions from video sequences to support activity understanding, workflow analysis, and safety applications.

Lightweight anomaly detector for time series data using Isolation Forest, designed for edge environments.

Real-time AI-powered vision system for detecting defects in manufacturing environments

Transcribes spoken audio into text using the Whisper speech recognition model, optimized for edge inference.

Real-time environmental sound classifier optimized for low-power edge devices using MobileNet v1.

Transcribes spoken audio into text using Wav2Vec2 for edge-based speech recognition applications.