Ev.lib
Embedded Computer Vision Library
EV Lib is a complete embedded vision software library based on deep learning and AI with functionalities for people, vehicle and object detection, identification tracking and 3D pose estimation.
EVLib is a highly optimized embedded vision software library that delivers powerful customizable AI and machine learning at the edge.
Using EV Lib’s runtime models and versatility, you can design and implement scalable, low power and low deployment cost AI-powered vision system without sacrificing performance.
Applications & Functions
EVLib offers a wide flexibility for specific object detection like food recognition, product catalogues recognition and numerous other applications. Pre-trained models can be easily extended for custom object detection and tracking, and can be customized to any computer vision application using scenario-specific data.
People
- People counting and tracking
- Soft Biometrics (gender, age group)
- First Person vision applications
- Safety monitoring (helmet, vest)
Vehicles
- Vehicle detection, counting and tracking
- Vehicle classification and recognition
- LPR
- Lane detection
Objects
- Custom object detection in 2D/3D
- Object Pose Estimation
- Classification
- Segmentation
People Detection and Soft Demographics
People counting, occupancy monitoring in confined or restricted areas, daily people traffic analysis and precious analytics such as gender and age group for various scenarios can be possible, regardless of area and hardware limitations.
Computer vision systems empowered with EV Lib can be installed virtually anywhere, guarantying GDPR compliancy and data security, while keeping power consumption and hardware cost to a minimum.
- Build a vision sensor (IoT device) with Computer Vision and Edge AI
- Anonymizer (Privacy-preserving) – Blurring people to remove sensitive data (GDPR)
- Cost reduction >80%
https://www.youtube.com/watch?v=4c1zNDlsklI
Objects and Vehicle Detection and Tracking
General and specialized vehicle detection, traffic monitoring, space and parking lot occupancy and analytics, interactive city and traffic lights, pedestrian or personnel safety, obstacle avoidance (e.g. potholes and speed bumps), vision aided infrastructure, all of the above are part of the modern smart city landscape. EV Lib can help to robustly analyze an urban scene, bringing AI-empowered vision exactly where it is most valuable.
https://www.youtube.com/watch?v=SQkGpzaUHEE
A complete embedded computer vision software library
Shopping with AI
Highly scalable, low power solution for automatic counting and identification of shopping cart contents and products.
• Real-time performance
• Low power consumption
• Low hardware cost
• Low-volume data campaigns
• Integrated solution: edge processing only
Logistics
• 2D/3D object detection
• Process automation in logistics and construction sites
• Bringing intelligence and automation in a totally new market
• Rated in the top 5 companies in insulation products globally
• Reducing cost of operations and resources >50%
Smart Space
• Camera-based real time parking space inspection
• Modular structure for scalability and adaptation to different use cases
• Covers indoor/outdoor, on-street/off-street parking scenarios
• Runs on any device and/or OS
• No need for digging and installing underground sensors.
Food Industry
• Custom 2D food detection and recognition
• Object detection /classification
• Reduce proc. time by 25% using our real-time recognition engine
How it works
Target Concept
Use existing EV-Lib functionalities
Deploy EV-Lib
Distribute first version | work from day-1
Optimize EV-Lib
Optimize models on target scenario and hardware | Continuous update and releases | Specialized models for every customer | Learn from new samples
Release Final Version
Check for criteria | Release final version | Lifelong learning and self adaptation
- Deploy fleet of devices in various architectures
- EV-Lib built-in mechanism for model updates
- Architectures for deployment
- Device level
- Local network level
- (Near Edge Device) NED
- Deploy on a device close to the data
- Monitor one or more architectures
- Workflow:
- Customer deploys EV-Lib models
- Training is performed at Irida Labs
- Dedicated server per customer
Markets
Commercial security, smart analytics, soft demographics and safety in the GDPR era
Computer vision based awareness in very small devices, making everyday life easy
Efficient traffic monitoring, utility, space, and pedestrian management for the urban infrastructure
Edge AI machine vision industrial automations, driving the 4th industrial revolution
Product detection, tracking and vision aided inventory for accurate warehouse management
Partners and Customers
Trusted by world-class leaders in the fields of embedded systems, digital cameras, semiconductors, manufacturing and telecommunications, we have built an ecosystem capable to holistically support even the most challenging computer vision applications.
Previous
Next
Would you like to know more?
Get in touch with our experts to learn more about the capabilities of EVLib and explore how AI and deep learning can take your product or system to the next level!