How 5758703/cv_pythonvue_tigerpro works
Tiger AI is a Flask/Vue multi-task vision and speech AI platform with an admin console and public portal. The primary workflow is an operator configuring a camera or analysis workflow, submitting it through the console, and receiving inference, tracking, alerts, and persisted results. The graph emphasizes documented AI scenarios, cross-camera MTMC, configurable video pipelines, and platform access control. Most feature routes and several integrations are represented from README capabilities and the tree; links without source-level evidence are omitted rather than inferred.
Main entry points: index.vue Admin console · pipeline_runtime.py Pipeline runtime · mtmc_engine.py Cross-camera MTMC · pipeline.py Pipeline API
Components 24
User interfaces
- Project portal
frontend/frontend_home/src/App.vue - Admin console
frontend/frontend_admin/src/views/ai/scenarios/index.vue - Pipeline editor
frontend/frontend_admin/src/views/ai/pipeline/index.vue
Platform access
- JWT and RBAC
backend/routes/auth.py - Identity models
backend/models/__init__.py - Model management
backend/routes/ai_model.py
AI inference
- Model scenarios
backend/services/model_scenarios.py - Scenario inference
backend/services/model_scenario_inference.py - Inference service
backend/inference.py - Speech transcription
backend/services/moss_mtd.py - Model training
backend/services/training.py
Video analytics
- Camera streams
backend/services/camera_stream.py - Cross-camera MTMC
backend/services/mtmc_engine.py - Local tracking
backend/services/mtmc_local_track.py - MTMC evidence
backend/models/mtmc.py - Alert rules
backend/services/alert_engine.py
Pipeline runtime
- Pipeline API
backend/routes/pipeline.py - DAG validation
backend/services/pipeline_schema.py - Pipeline runtime
backend/services/pipeline_runtime.py - MTMC bridge
backend/services/pipeline_mtmc.py - Pipeline records
backend/models/pipeline.py
Other components
- Operator
- MQTT broker
- Webhook target
Connections 29
- Operator → Project portal: opens portal
README.md - Operator → Admin console: uses console
README.md - Project portal → Admin console: links to
README.md - Admin console → JWT and RBAC: authenticates
- JWT and RBAC → Identity models: checks identity
backend/routes/auth.py - Admin console → Model management: manages models
- Admin console → Model scenarios: selects scenario
- Model scenarios → Scenario inference: invokes
- Scenario inference → Inference service: runs inference
- Admin console → Speech transcription: submits media
- Admin console → Model training: manages training
- Admin console → Camera streams: views streams
- Camera streams → Cross-camera MTMC: provides streams
backend/services/camera_stream.py - Cross-camera MTMC → Local tracking: tracks locally
backend/services/mtmc_engine.py - Cross-camera MTMC → Inference service: requests detections
backend/services/mtmc_engine.py - Cross-camera MTMC → MTMC evidence: persists evidence
backend/services/mtmc_engine.py - Admin console → Pipeline editor: configures pipeline
- Pipeline editor → Pipeline API: calls API
- Pipeline API → DAG validation: validates DAG
backend/routes/pipeline.py - Pipeline API → Pipeline records: stores definitions
backend/routes/pipeline.py - Pipeline API → Pipeline runtime: starts runtime
backend/routes/pipeline.py - Pipeline runtime → Camera streams: reads streams
backend/services/pipeline_runtime.py - Pipeline runtime → Inference service: runs detection
backend/services/pipeline_runtime.py - Pipeline runtime → Alert rules: evaluates rules
backend/services/pipeline_runtime.py - Pipeline runtime → Pipeline records: records runs
backend/services/pipeline_runtime.py - Pipeline runtime → MTMC bridge: bridges MTMC
backend/services/pipeline_runtime.py - MTMC bridge → Cross-camera MTMC: attaches session
backend/services/pipeline_mtmc.py - Pipeline runtime → MQTT broker: publishes events
backend/services/pipeline_runtime.py - Pipeline runtime → Webhook target: sends events
backend/services/pipeline_runtime.py