Upload v2/check.py with huggingface_hub

v2/check.py

@@ -0,0 +1,104 @@
+from pathlib import Path
+import trimesh
+import numpy as np
+
+
+def check_head_model(model_path: str | Path) -> dict:
+    """Validate a head model mesh."""
+    model_path = Path(model_path)
+    if not model_path.exists():
+        return {"status": "error", "message": f"Model not found: {model_path}"}
+
+    try:
+        scene = trimesh.load(str(model_path))
+    except Exception as e:
+        return {"status": "error", "message": f"Failed to load model: {e}"}
+
+    results = []
+    if isinstance(scene, trimesh.Scene):
+        for name, g in scene.geometry.items():
+            if isinstance(g, trimesh.Trimesh):
+                centroid = g.vertices.mean(0)
+                y_min, y_max = g.vertices[:, 1].min(), g.vertices[:, 1].max()
+                results.append({
+                    "mesh": name,
+                    "centroid": centroid.round(3).tolist(),
+                    "y_range": [float(y_min), float(y_max)]
+                })
+    else:
+        centroid = scene.vertices.mean(0)
+        y_min, y_max = scene.vertices[:, 1].min(), scene.vertices[:, 1].max()
+        results.append({
+            "mesh": model_path.stem,
+            "centroid": centroid.round(3).tolist(),
+            "y_range": [float(y_min), float(y_max)]
+        })
+
+    return {"status": "ok", "results": results}
+
+
+def check_body_model(model_path: str | Path) -> dict:
+    """Validate a body model mesh with anatomical checks."""
+    model_path = Path(model_path)
+    if not model_path.exists():
+        return {"status": "error", "message": f"Model not found: {model_path}"}
+
+    try:
+        scene = trimesh.load(str(model_path))
+    except Exception as e:
+        return {"status": "error", "message": f"Failed to load model: {e}"}
+
+    if isinstance(scene, trimesh.Scene):
+        body = trimesh.util.concatenate(list(scene.geometry.values()))
+    else:
+        body = scene
+
+    v = body.vertices
+    y_min, y_max = v[:, 1].min(), v[:, 1].max()
+
+    checks = [
+        ('feet', 0.01),
+        ('hips', 0.50),
+        ('chest', 0.65),
+        ('neck', 0.75),
+        ('chin', 0.87),
+        ('crown', 0.99)
+    ]
+
+    results = []
+    for name, frac in checks:
+        y = y_min + (y_max - y_min) * frac
+        pts = v[(v[:, 1] > y - 0.03) & (v[:, 1] < y + 0.03)]
+        if len(pts) < 3:
+            continue
+        results.append({
+            "location": name,
+            "fraction": frac,
+            "x_mean": float(pts[:, 0].mean()),
+            "x_std": float(pts[:, 0].std()),
+            "z_mean": float(pts[:, 2].mean()),
+            "z_std": float(pts[:, 2].std()),
+            "points": int(len(pts))
+        })
+
+    return {"status": "ok", "y_range": [float(y_min), float(y_max)], "results": results}
+
+
+def check_model(model_path: str | Path, model_type: str = "auto") -> dict:
+    """Check a model, auto-detecting type if needed."""
+    model_path = Path(model_path)
+
+    if model_type == "auto":
+        if "head" in model_path.name.lower():
+            model_type = "head"
+        elif "body" in model_path.name.lower():
+            model_type = "body"
+        else:
+            return {"status": "error", "message": "Could not auto-detect model type. Use --model-type to specify (head/body)"}
+
+    if model_type == "head":
+        return check_head_model(model_path)
+    elif model_type == "body":
+        return check_body_model(model_path)
+    else:
+        return {"status": "error", "message": f"Unknown model type: {model_type}"}