Viso Suite feature index
Viso Suite provides a uniquely powerful end to end platform for computer vision development, deployment, monitoring and scaling. Explore the list of features.
Workspace
Core features
- Cloud-native workspace
- Fully-managed infrastructure
- Over 15 end to end products
- Seamless moving between tools
- Secure in-browser access
Administration
- 24×5 support
- Help desk, email support
- Tailored service levels
- Usage stats and reports
- Support access
Professional services
- Requirement engineering
- Technical solution review
- Full solution building
- Industry expertise, advisory
- Annotation services
- Model training services
Infrastructure
- Model-driven viso architecture
- Managed infrastructure
- On premises deployment
- Auto-scaling
- Load balancing
Multitenancy
- Multiple workspaces
- Fully isolated tenants
- Join multiple workspaces
- Seamless workspace switching
Workspace customization
- Company logo
- Color themes
- Dark/Light mode
- Translations
Workspace center
- Customized home view
- Modular widgets
- Systems health
- Status, usage report
Data collection
Core features
- Video cloud
- Centralized cloud storage
- Edge data collection
- Collect data from connected devices
- Multi-camera recording
- Fleet Edge collection
- Distributed edge collection
- Reliable data delivery
- Store and manage video assets at scale
Automated collection
- Automated recording
- Image and video data
- Scheduling
- Continuous collection
- Custom, conditional logic
- Multi-device support
- Offline recording
Video recording
- Remote recording
- One-click capturing
- Recording tags
- HD recording
- Automatic processing
Data organization
- Video manager
- Advanced filters
- Data tags
- Built-in video player
- Commenting on files
Data integration
- Video data import
- MP4
- AVI
- MOV
- Drag-and-drop upload
- Large files
- Video transcoding
- Import hosted videos
- Cloud import
Data administration and security
- Secure central data repository
- Zero Trust Application Access
- Data access control
- Log user activity
- Full data visibility and management
- 256-bit encryption at rest
- TLS, SSL encryption during transit
- Manual session timeouts
Dataset curation
- Dataset explorer
- Analyze and visualize datasets
- Import and export datasets
- Dynamic search, filter, sort
- Merge datasets
- Image uniqueness analysis
- Find duplicates/near-duplicates
- Remove duplicate objects
- Find unique/representative samples
- Annotation mistake analysis
- Find and fix label mistakes
- Identify missed objects
- Incorrect annotations
- Localization issues
- Analyze bad samples
- Identify data collection scenarios
- Strengthen data variability
- Reduce dataset bias
- Identify training data gaps
Data annotation
Core features
- Image labeling
- Video annotations
- Installation-free
- Import pre-annotations
- Basic/Advanced usage modes
- Automatic authentication
Annotation tools
- 2D bounding box, rectangle
- 3D object annotation, cuboids
- Polylines, lines, splines
- Point annotation, ellipses
- Keypoint, landmarks, keypoint naming
- Image classification, image captioning
- Semantic segmentation
- Occlusion, background/foreground layering
- Attribute annotation, tag annotation
- Shape grouping
- Action shortcuts
Administration
- Manage projects
- Annotation tasks
- Task management
- Quality management
- Role based access control
- Annotation analytics
Automated annotation
- Automatic annotation
- ML-assisted labeling (semi-automatic)
- Object predictions, detectors, interactors
- Backpropagating refinement scheme (BRS)
- Polygon generation, Mask RCNN, Faster RCNN
- Deep Extreme Cut
- High-Resolution Net
- Inside-Outside-Guidance
- Deep learning based trackers
- Active learning
Video annotation
- Video keyframes, sequences
- Video annotation timeline
- Playback speed variation
- Automatic object tracking across frames
- Movement tracking
- Object codification, object ID
- Video FPS limiting, frame skipping
- SD, HD, Ultra HD 4K videos
- Cross-camera sources (IP, CCTV, USB, GoPro, Drones)
Annotation analysis
- Assess annotation quality
- Identify annotation gaps
- Measure KPIs
- Annotation distribution
Formats
- MS COCO object detection
- PASCAL VOC segmentation masks
- YOLO
- Datumaro
- LabelMe
- TFRecord
- MOT and MOTS
- Tensorflow object detection API
- ImageNet
- CamVid
- WIDER Face
- VGG Face2
- Market-1501
- ICDAR 13/15
- Open Images
- Cityscapes
- KITTI
- LFW
- CVAT
Annotation workforce
- External/internal annotators
- Team collaboration tools
- Assign annotation jobs
- Labeling services
- Limited guest roles
Model training
Core features
- Manage and train AI models
- Pre-made training pipelines
- Custom training pipelines
- Dedicated GPU, CPU resources
- Import custom algorithms
- Train and retrain algorithms
Augmentation capabilities
- Add noise, invert, compression
- Cutout, dropout, replace
- Blend alpha, mask, grid
- Gaussian blur, motion blur
- Color, brightness, saturation, hue
- Contrast, gamma, histogram
- Convolve, sharpen
- Edge detector, embossFlipping, mirroring, rotation
- Shifting, scaling, distortion, jigsaw
- Auto-contrast, enhancing, solarize
- Corruption: defocus, zoom-blur, pixelate
- Pooling: average, min, max median
- Segmentation: superpixels, voronoi
- Resize, cropping, crop and pad
- Weather: clouds, fog, frost, snow, rain
- Combine augmenters
- Sequential combination
- Randomize
- Shuffle, p-percent
AI model manager
- AI model library
- Framework independent
- Add pre-trained models
- Add custom models
- Version control
- Dependencies
- Tags and filters
- Conflict detection
- Built-in access control
Training environments
- Automated infrastructure
- Multiple training environments
- Auto-scaling infrastructure
- Automatic authentication
- One click creation
- Pause and resume
Model training evaluation
- Evaluate model performance
- Model prediction evaluation
- Regressions
- Classifications
- Detections
- Instance segmentation
- Semantic segmentation
- Classification reports
- Confusion matrices
- Precision-Recall curves
- Model accuracy analysis
- Cluster false positive/low accuracy
- Render labels on sample images
- Interactive plots
- Anomalies: Find anomalous training data
- Mistakes: Detect incorrect prediction patterns
- Hardness: Mine hard samples
- Unsupervised data pre-annotation
- Identify scenarios that require more training data
- Continuous MLOps
- Strengthen dataset robustness
- Reduce annotation mistakes
- Data collection strategies
Ready-made environments
- Jupyter environments
- Pytorch
- TensorFlow 2
- TensorFlow Lite
- OpenVINO
- JupyterLab Base
- Conda virtual environments
- Custom Viso JupyterLab Images
- Pillow
- lxml
- Cython
- Jupyter
- Matplotlib
- Pandas
- OpenCV Python
- Training and retraining
- OpenPifPaf
- DeepFace
- MediaPipe
- Keras
- Caffe
Pre-trained algorithms
- CPU, GPU, VPU, TPU inferencing
- Object Detection
- People Detection
- Animal Detection
- Image Classification
- Visual Inspection
- Clustering
- Visual Search
- Keypoint Detection
- Pose Estimation
- Gaze Estimation
- Hand Pose Estimation
- Object Pose Estimation
- Tracking
- Object Tracking
- Object Counting
- People Counting
- Dwell Time Tracking
- Activity Analysis
- Object Segmentation
- Motion Analysis
- Activity and motion heatmap
- Fall detection
- Posture recognition
- Anomaly detection
- Face Analysis
- Face Detection
- Face Recognition
- Emotion Analysis
- Gender, age analysis
Data augmentation
- ML image augmentation
- Augmentation environments
- Debug augmented images
- Image load and augment
- Multicore Augmentation
- Augment Keypoints (Landmarks)
- Augment Bounding Boxes
- Augment Polygons
- Augment Line Strings
- Augment Heatmaps
- Augment Segmentation Maps
- Probability distribution parameters
- Stochastic & deterministic Augmentation
- Copy-Paste augmentation
- Multiple Augmentation Sequences
Application development
Core features
- Computer vision Applications
- Build flows with modules
- Swap modules and AI models
- Hardware-software independent
- Cross-platform portability
- Scalable by design
Processing
- High-performance AI vision
- Low latency edge computing
- Cloud, hybrid: Virtualized edge
- Process on CPU, GPU, TPU, VPU
- Stack processors (e.g. multi-GPU)
- Combine processors (e.g. CPU+GPU)
- Frame buffer, frame-store
- Multi-threading, parallel processing
- Batch processing
- Dynamic buffering
- Workload orchestration
- Edge load balancing
- Offline support
- Privacy-preserving AI
System integration
- Custom logic flows
- Code snippets (JS, Python)
- Custom Module SDK
- Add docker containers
- Open API support
- HTTP and MQTT
- External databases
Applications
- Application Library
- Advanced filters, tags
- Template manager
- Custom templates
- Clone applications
- Export applications
- Import applications
- Auto dependency manager
- Cross-workspace sharing
- Semantic version control
- One-click roll backs
- User activity logs
Modules
- Module library
- Pre-built Viso modules
- Module auto-installer
- Custom modules (SDK)
- Container manager
- Custom containers
- Version control
- Update management
- Visual configuration
Input
- Live camera: IP, CCTV, USB, etc.
- Web stream: WebRTC, RTSP, etc.
- Virtual cameras
- Externally hosted videos (Youtube, Vimeo, etc.)
- External data storage, s3 etc.
- Video data manager
- Multi stream input
- Image processing
Computer vision
- Drop-down AI model select
- AI model library integration
- One-click swap AI models
- Key Viso Modules
- Object Detection
- Image Segmentation
- Video Recognition
- Image Classification
- Group Key Points
- Movement Analysis
- Posture Analysis
- Object Counting
- Fall Detection
- Object Tracking
- Motion, visual flow
- 3D Keypoint detection
- Face analysis
- Face recognition
- Text recognition (OCR)
- Barcode reader
Image pre-processing
- Image coloring, sharpening
- Distortion correction, de-warping
- Region of Interest (ROI)
- Multi-regions of Interest
- Area including, excluding
- Complex ROI polygons
- Image grid sections
Application Editor
- Collaborative development
- Visual development editor
- Drag-and-drop flows
- Visual configuration
Deployment
Core features
- Automated deployment
- Run AI vision at the edge
- Unified endpoint manager
- Automated deployment
- Scale distributed systems
- Isolated, hardened endpoints
- Edge AI Computer Vision
Device management
- Edge devices
- Virtual devices (hybrid)
- Simple management
- Device attributes
- Device group tags
- Advanced filters
- Device authentication
- Ruggedized Edge
Computer Vision OS
- Computer vision optimized OS
- Active and passive protection
- Remote updating
- Create and manage OS images
- Automated OS installation
Device fleets
- Fleet management
- Multi-platform fleet
- Mass device enrollment
- Bulk deployment
- Separated environments
- Automated diagnosis
- Scheduled health scans
Automated deployment
- Deployment profiles
- One-click deployment
- Release orchestration
- Automated pipelines
- Impact analysis
- Fail-safe automation
- Roll-back tools
- OTA updating
- Disconnected Edge
Device lifecycle
- Hardened Operating System
- Multi-architecture OS support
- Device type manager
- No-code enrollment
- Automated provisioning
- Remote enrollment
- Disconnected Edge
- Device status, events
- Securely retire devices
Security
- Cloud device twin
- Active edge defender
- Secure erasure
- Theft protection, remote wipe
- Protected app and model
- Identity management
- Encryption at rest and in transit
- Authentication, authorization
- Multi-level access control
Edge Machine Learning
- Physical and virtual devices
- Local vision processing
- Cross-platform Edge AI
- Generic computing devices x86
- Intel NUC amd64
- Up Board amd64
- Server and MEC support
- AWS AMI
- Processor agnostic
- Central Processing Unit (CPU)
- Graphics Processing Unit (GPU)
- Vision Processing Unit (VPU)
- Tensor Processing Unit (TPU)
- Multi-Processor computing
Manage
- Remote stream view
- Output preview stream
- Multi stream grid
- Secure preview timeout
- Remote control, reboot
- Live analytics dashboard
- Real-time and historical
- Drill-down, dynamic charts
- Hardware: CPU, GPU, threads
- Network: Bandwidth, bit-rate
- Storage, containers
- Deployment analytics
- Edge Load Balancing
- Configuration Manager
- Zero trust SSH terminal
- Automated health scan
Connectivity
Core features
- Cloud-Edge distributed AI vision
- Fully managed infrastructure
- Real-time applications
- High-performance time series
- Elastic, robust auto-scaling
Distributed Edge
- Always-on computer vision
- Edge device fleet integration
- On-device AI processing
- Edge-cloud communication
- Low-latency processing
Time series
- Managed time series database
- Scalable, large data volumes
- Real-time events
- Critical stateful data
- Custom metrics
Security and privacy
- Private data processing
- Anonymous metrics
- Secure 2-way communication
- Enforced SSL, TLS encryption
- Secure Edge Tunneling (VPN)
- Cloud-based authentication
- Managed users, permissions
- Edge on-disk encryption
Edge Gateway
- Integrated Edge Gateway
- Automated initialization
- Synchronization
- Message Broker
- MQTT V3.1, V5.0 Support
Robust connectivity
- Full streaming support
- Store and forward
- Structured, unstructured data
- Operate online or offline
- Robust delivery guarantees
- Auto scheduler
- Clock-drift adjustment
- Analytics tool connectors
Edge agent
- Automated collection
- Custom Formatting
- Metrics aggregation
- Writing edge metrics
- Flexible storage, retention
- In-memory buffering
Monitoring
Core features
- Unified observability
- End-to-end stack
- Application monitoring
- Real-time analytics
- Historical data analytics
Fully managed analytics
- No servers to manage
- Auto-scaling infrastructure
- High-availability
- Security built-in
Team collaboration
- Shared dashboards
- Real-time view and editing
- Version control
- Team troubleshooting
- External dashboard sharing
Unified observability
- Application metrics monitoring
- Federated aggregation across devices
- Edge device dashboard
- Health monitoring dashboard
Data integration
- Automated data connectors
- Live data streaming
- Historical data synchronization
- Dynamic data filters
- Custom data query builder
- Alert notifications
Security
- User authentication, authorization
- Automated database connectors
- Auto-encryption of data at rest
- Data in transit encryption (TLS)
Data visualization
- Unlimited dashboards
- Pre-built dashboards
- Ready-made widgets
- Graphs
- Bars
- Gauges
- Heatmaps
- Maps
- Tables
- Cards
- Custom widgets
- Layer, segment data points
- Interactive panels
- Drill-down explore data
Maintain
Core features
- Application maintenance
- Configure applications at scale
- MLOps capabilities
Diagnose and troubleshoot
- Notification center
- System alerts
- Client-less remote control
- Remote system health scanning
- Automated issue detection
- Identify defect hardware
- Remote command execution
- Monitor endpoint resources
- Disaster Recovery
Collaboration
- User authentication and authorization
- Seamless team collaboration
- User activity log
Configuration
- Configuration profiles
- Manage configurationsVisual
- configuration editor
- Set processing parameters
- One-click deployment
- Config version control
- Configuration recovery, roll-back
- Draft changes
Application maintenance
- Maintain applications in production
- Release application updates
- OTA update delivery
- Staged roll-outs
MLOps capabilities
- Continuous model training
- Model evaluation and optimization
- Model catalog, versioning
- Automated model deployment
- Swap and replace models
Security and privacy
Core features
- End-to-end security stack
- Centralized data/asset protection
- Zero Trust Security
- Distributed Edge Security
- AWS infrastructure security
Compliance
- GDPR compliant
- ISO27001 compliant
- SOC 2 Type 2 compliant
Unified security
- User management
- Invitation manager
- RBAC Authorization
- Identity Management (users, devices)
- Custom user roles, teams
- Granular access restriction
- Invitation management
- Multi-factor authentication (MFA)
- Log and trace activities
- Active Edge defender scanning
- Loss protection, auto wiping
- Active model protection
- Secure remote edge access
- SSH Zero Trust Security
Policies and compliance
- Password policies
- Session policies
- Pre-built password policies
- Enforce password strength
- Enforcing MFA
Operational security at viso.ai
- Adherence to control frameworks
- Risk management program
- Background checks of employees
- Security and privacy awareness training
- Least privilege, separation of duties
- Need-to-know access for information
- Secure software development lifecycle
- Separation of environments
- Monitoring and alerting for events
- Regular control assessments
- Regular vulnerability scans
- Comprehensive penetration tests
- Risk-based prioritization of vulnerabilities
- Malicious software protection
- Data loss prevention
- Incident management planning and testing
- Business continuity, disaster recovery
- Capacity and resource planning
- Secure corporate networks, SDP
Data security
- Encryption at rest (AES-256)
- Encryption in transit (SHA-256, RSA 2048)
- Secure virtual private clouds (VPC)
- Secure virtual private networks (VPN)
- Server hardening at the Edge
- PFS encryption
Application security
- Zero-Trust Application Access
- Multi-layer security
- Dependency scanning
- Container vulnerability scans
- API authentication
- Authentication policies
- Data validation
- Encryption at rest and in transit
- Audit logging
- Hosted platform security
- Infrastructure as code (IaC)
- Logical content separation
- Man-in-the-middle prevention
- Role-based authorization
- Secure credential storage
- Session policies
- Download prevention
Privacy
- GDPR compliant computer vision
- No video transfer/storage (Edge AI)
- Fully autonomous applications
- Zero Trust Live View restriction
- Anonymization
- Metrics output
- Blurring, Pixelation
- Area blocking
- Disable visual output
- Needs-to-see permission access
- Separate environments
- Stream auto timeout
- Auditable, transparent data flows