phase cv 0

gpu/models/inference_contract.py

@@ -0,0 +1,112 @@
+"""
+Pydantic Models - GENERATED FILE
+
+Do not edit directly. Regenerate using modelgen.
+"""
+
+from datetime import datetime
+from enum import Enum
+from typing import Any, Dict, List, Optional
+from uuid import UUID
+
+from pydantic import BaseModel, Field
+
+class DetectRequest(BaseModel):
+    """Request body for object detection."""
+    image: str
+    model: Optional[str] = None
+    confidence: Optional[float] = None
+    target_classes: Optional[List[str]] = None
+
+class BBox(BaseModel):
+    """A detected bounding box."""
+    x: int
+    y: int
+    w: int
+    h: int
+    confidence: float
+    label: str
+
+class DetectResponse(BaseModel):
+    """Response from object detection."""
+    detections: List[BBox] = Field(default_factory=list)
+
+class OCRRequest(BaseModel):
+    """Request body for OCR."""
+    image: str
+    languages: Optional[List[str]] = None
+
+class OCRTextResult(BaseModel):
+    """A single OCR text extraction result."""
+    text: str
+    confidence: float
+    bbox: List[int] = Field(default_factory=list)
+
+class OCRResponse(BaseModel):
+    """Response from OCR."""
+    results: List[OCRTextResult] = Field(default_factory=list)
+
+class PreprocessRequest(BaseModel):
+    """Request body for image preprocessing."""
+    image: str
+    binarize: bool = False
+    deskew: bool = False
+    contrast: bool = True
+
+class PreprocessResponse(BaseModel):
+    """Response from preprocessing."""
+    image: str
+
+class VLMRequest(BaseModel):
+    """Request body for visual language model query."""
+    image: str
+    prompt: str
+    model: Optional[str] = None
+
+class VLMResponse(BaseModel):
+    """Response from VLM."""
+    brand: str
+    confidence: float
+    reasoning: str
+
+class AnalyzeRegionsRequest(BaseModel):
+    """Request body for CV region analysis."""
+    image: str
+    edge_canny_low: int = 50
+    edge_canny_high: int = 150
+    edge_hough_threshold: int = 80
+    edge_hough_min_length: int = 100
+    edge_hough_max_gap: int = 10
+    edge_pair_max_distance: int = 200
+    edge_pair_min_distance: int = 15
+
+class RegionBox(BaseModel):
+    """A candidate region from CV analysis."""
+    x: int
+    y: int
+    w: int
+    h: int
+    confidence: float
+    label: str
+
+class AnalyzeRegionsResponse(BaseModel):
+    """Response from CV region analysis."""
+    regions: List[RegionBox] = Field(default_factory=list)
+
+class AnalyzeRegionsDebugResponse(BaseModel):
+    """Response from CV region analysis with debug overlays."""
+    regions: List[RegionBox] = Field(default_factory=list)
+    edge_overlay_b64: str = ""
+    lines_overlay_b64: str = ""
+    horizontal_count: int = 0
+    pair_count: int = 0
+
+class ConfigUpdate(BaseModel):
+    """Request body for updating server configuration."""
+    device: Optional[str] = None
+    yolo_model: Optional[str] = None
+    yolo_confidence: Optional[float] = None
+    vram_budget_mb: Optional[int] = None
+    strategy: Optional[str] = None
+    ocr_languages: Optional[List[str]] = None
+    ocr_min_confidence: Optional[float] = None