How parshotam94/cv_assurance works
VisionTrust Assurance is an offline computer-vision integrity and provenance system. A local user operates the dashboard or API to assess datasets and models, check distribution drift, and run inference; inference records are signed and can be verified against tampering or replay. Findings support risk disposition, audit and compliance, and PDF/JSON/CSV dossiers. The README documents these capability areas, while sampled source confirms application routing, database/repository behavior, and inference verification flows; some internal wiring is therefore left unspecified.
Main entry points: main.py FastAPI app · inference.py Inference API · datasets.py Dataset API · models.py Model API
Components 20
Local interface
- FastAPI app
backend/app/main.py - Dataset API
backend/app/api/datasets.py - Model API
backend/app/api/models.py - Distribution API
backend/app/api/distribution.py - Inference API
backend/app/api/inference.py - Offline dashboard
frontend/js/api.js
Asset assurance
- Dataset analysis
backend/app/dataset - Model analysis
backend/app/model - Distribution analysis
backend/app/distribution
Provenance and audit
- Provenance verification
backend/app/inference - Signing keys
backend/app/core/signatures.py - Audit ledger
backend/app/core/audit.py
Risk and reporting
- Risk engine
backend/app/core/risk_engine.py - Findings API
backend/app/api/findings.py - Compliance API
backend/app/api/compliance.py - Reports API
backend/app/api/reports.py - Dossier generation
backend/app/reports - Repository layer
backend/app/database/repository.py - SQLite database
backend/app/database/database.py
Other components
- Local user
Connections 19
- Local user → Offline dashboard: uses
- Offline dashboard → FastAPI app: API requests
- Local user → FastAPI app: API requests
- FastAPI app → Dataset API: routes
- FastAPI app → Model API: routes
- FastAPI app → Distribution API: routes
- FastAPI app → Inference API: routes
- FastAPI app → Findings API: routes
- FastAPI app → Compliance API: routes
- FastAPI app → Reports API: routes
- Dataset API → Dataset analysis: analyzes
- Model API → Model analysis: inspects
- Distribution API → Distribution analysis: detects drift
- Inference API → Provenance verification: verifies records
- Inference API → Signing keys: signs records
- Inference API → Repository layer: stores records
- Findings API → Risk engine: scores findings
- Reports API → Dossier generation: generates dossiers
- Repository layer → SQLite database: reads and writes