Skip to main content

Settings Glossary

This glossary provides definitions for all settings across Anava Skills, Detections, and Groups.

A​

Active (Detection)​

Type: Boolean Default: true Location: Detection Settings

Whether the Detection is enabled. When false, the Detection does not process events.

Active Monitoring​

Type: Object Location: Detection Settings

Configuration for continuous frame capture and temporal analysis. See Learning Mode.

Analysis Schedule​

Type: String Default: "24/7" Location: Detection Settings

When the Detection should be active. Values:

  • 24/7 - Always active
  • business_hours - Camera-defined business hours
  • Custom cron expression

Author​

Type: String Location: Skill Settings

Creator of the skill. For documentation and attribution.

B​

Batch Size​

Type: Number (1-10) Default: 1 Location: Active Monitoring Settings

Number of frames sent together for analysis. Higher values provide more context but increase latency.

C​

Category​

Type: String Location: Skill Settings

Organizational category for the skill. Common values:

  • Security
  • Safety
  • Operations
  • Compliance

Confidence Threshold​

Type: Number (0-100) Default: 0 Location: Detection Settings

Minimum confidence percentage required to emit ONVIF events. Higher values reduce false positives but may miss valid Detections.

D​

Description​

Type: String Location: Skill, Profile, Group Settings

Human-readable description of purpose and function.

Draw Bounding Box​

Type: Boolean Default: false Location: Object Settings

When true, detected objects are highlighted with bounding boxes in stored images.

E​

Enabled (Object)​

Type: Boolean Default: true Location: Object Settings

Whether to detect this object type. Disabled objects are not analyzed.

Enabled (Question)​

Type: Boolean Default: true Location: Question Settings

Whether this question is active. Disabled questions are not answered.

Enabled (Active Monitoring)​

Type: Boolean Default: false Location: Active Monitoring Settings

Whether continuous frame capture is enabled for the profile.

F​

Full Analysis Model​

Type: String Default: (system default) Location: Detection Settings

AI model used for full analysis phase. Leave empty for system default.

Full Analysis System Prompt​

Type: String Location: Skill Settings

System context provided to AI for full analysis. Establishes role and guidelines.

Full Analysis User Prompt​

Type: String Location: Skill Settings

Instructions for full analysis. Defines what AI should analyze and how.

G​

Group ID​

Type: String Location: Group Settings

Unique identifier for the group. Auto-generated.

I​

ID (Trigger)​

Type: String Location: Trigger Settings

Specific scenario ID for perimeter triggers. References AOAS scenario.

Interval Ms​

Type: Number (1000-60000) Default: 5000 Location: Active Monitoring Settings

Milliseconds between frame captures during active monitoring.

M​

Max Duration Sec​

Type: Number (30-600) Default: 60 Location: Active Monitoring Settings

Maximum duration in seconds for active monitoring window.

Max Images​

Type: Number (5-50) Default: 10 Location: Active Monitoring Settings

Maximum number of images to capture during active monitoring.

N​

Name​

Type: String Location: Skill, Profile, Group, Object, Question Settings

Unique identifier. Used in references and displays.

O​

Object​

Type: Array Location: Skill Settings

List of objects to detect. Each object has its own configuration.

P​

Port​

Type: Number (1-8) Location: Trigger Settings

I/O port number for DigitalInput or Manual triggers.

Pre-Filter Criteria​

Type: String Location: Skill Settings

Description of what makes an image worth analyzing. Guides AI pre-filtering.

Pre-Filter Model​

Type: String Default: (system default) Location: Detection Settings

AI model used for pre-filter phase. Typically faster/cheaper model.

Pre-Filter System Prompt​

Type: String Location: Skill Settings

System context for pre-filter analysis.

Pre-Filter User Prompt​

Type: String Location: Skill Settings

Instructions for pre-filter. Determines if image proceeds to full analysis.

Profile (Trigger)​

Type: String Location: Trigger Settings

AOAS profile or scenario name for Object/Perimeter triggers. Refers to the camera-side analytics profile, not the Anava Detection.

Push Notification​

Type: Boolean Default: false Location: Object, Question Settings

When true, a Detection triggers a mobile push notification.

Q​

Question​

Type: Array Location: Skill Settings

List of questions for AI to answer about each image.

Question ID​

Type: Number Location: Question Settings

Unique identifier for the question within the skill.

Question Text​

Type: String Location: Question Settings

The question the AI will answer about the image.

Question Type​

Type: Enum Location: Question Settings

Data type of the answer:

  • bool - True/false
  • int - Integer number
  • string - Free text
  • set - Multiple choice
  • varchar(50) - Short text
  • varchar(500) - Long text

R​

Rapid Eligible​

Type: Boolean Default: false Location: Object, Question Settings

When true, confident pre-filter detection can skip full analysis.

Resolution Profile​

Type: Enum Default: BALANCED Location: Active Monitoring Settings

Image capture resolution:

  • TINY - 228×128
  • LOW - 455×256
  • BALANCED - 640×360
  • HIGH - 854×480
  • HD_720 - 1280×720
  • FULL_HD_1080 - 1920×1080
  • ULTRA - 2560×1440
  • CUSTOM - User-defined

Response Criteria​

Type: String Location: Skill Settings

Expected response format and structure for AI output.

S​

Skill ID​

Type: String Location: Detection Settings

Reference to the Skill this Detection uses for analysis.

Stateful​

Type: Boolean Default: false Location: Object, Question Settings

When true, generates ONVIF true/false events for VMS integration.

Status (Group)​

Type: Enum Default: active Location: Group Settings

Group status:

  • active - Group is operational
  • inactive - Group is disabled

T​

Tags​

Type: Array of Strings Location: Group Settings

Organizational tags for grouping and filtering.

Talkdown Buffer Ms​

Type: Number Default: 30000 Location: Detection Settings

Minimum milliseconds between TTS announcements. Prevents rapid repeat.

Talkdown Enabled​

Type: Boolean Default: false Location: Object, Question Settings

When true, a Detection can trigger text-to-speech announcement.

Talkdown Guidance​

Type: String Location: Object, Question Settings

Instructions for generating TTS message content.

Talkdown Priority​

Type: Number (1-10) Default: 5 Location: Object, Question Settings

Priority ordering when multiple TTS events occur. Higher number = higher priority.

Talkdown Rule​

Type: Object Location: Question Settings

Condition that must be met to trigger TTS. Contains operator and value.

Talkdown Style​

Type: String Location: Object, Question Settings

Voice tone and style for TTS announcement.

Temporal Prompt Template​

Type: String Location: Skill Settings

Template for temporal analysis across multiple frames during Active Monitoring.

Trigger​

Type: Object Location: Detection Settings

Event source configuration. See Trigger Types.

Trigger Deep Analysis​

Type: Boolean Default: false Location: Object, Question Settings

When true, a Detection always runs full analysis regardless of pre-filter.

TTS Config​

Type: Object Location: Detection Settings

Text-to-speech configuration including model and voice.

TTS Model​

Type: String Location: TTS Config

TTS generation model.

TTS Voice​

Type: String Default: Kore Location: TTS Config

Voice selection for announcements:

  • Kore - Authoritative, clear
  • Charon - Calm, professional
  • Aoede - Friendly, warm
  • Puck - Energetic
  • Fenrir - Deep, commanding

Type (Trigger)​

Type: Enum Location: Trigger Settings

Trigger source:

  • None - Detection disabled
  • Manual - Virtual input
  • DigitalInput - Physical I/O
  • Schedule - Time-based
  • Motion - Legacy VMD
  • VMD4 - AXIS VMD4 motion detection
  • Object - AOAS detection (legacy)
  • ObjectAnalytics - AXIS Object Analytics
  • Perimeter - AOAS line/zone
  • PerimeterDefender - Advanced perimeter
  • Pulse - Camera-based interval scheduling

U​

Use Single​

Type: Boolean Default: false Location: Active Monitoring Settings

When true, uses single frame mode instead of batch.

V​

Version​

Type: String Location: Skill, Group Settings

Version identifier for tracking changes.

View Area​

Type: Number (1-8) Default: 1 Location: Detection Settings

Camera stream/view to capture. For multi-sensor cameras or PTZ presets.