{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6b701f06",
   "metadata": {},
   "source": [
    "# 09 · Visión más allá de la clasificación: segmentación, detección y similitud\n",
    "\n",
    "El notebook 04 respondía una pregunta por imagen: *¿qué dígito es este?* La mayoría de los sistemas reales de visión artificial necesitan respuestas más ricas:\n",
    "\n",
    "| Tarea | Salida | Usos de ejemplo |\n",
    "|---|---|---|\n",
    "| clasificación (notebook 04) | una etiqueta por imagen | etiquetado de fotos, control de calidad |\n",
    "| **segmentación semántica** | una etiqueta por **píxel** | imagen médica, usos del suelo desde satélite, \"zona transitable\" en conducción autónoma |\n",
    "| **detección de objetos** | una **caja + etiqueta** por objeto | conteo, seguimiento, conducción autónoma, estanterías de tiendas |\n",
    "| **aprendizaje métrico** | un **embedding** donde la distancia significa similitud | verificación facial, búsqueda de imágenes, reconocer clases nunca vistas en el entrenamiento |\n",
    "\n",
    "Construimos una versión pequeña y honesta de cada una, todo en una CPU en pocos minutos. Las imágenes se generan en el notebook: escenas sencillas de círculos, cuadrados y triángulos, con máscaras y cajas de referencia exactas. Así cada métrica es exacta y podemos mirar dentro de los modelos sin descargar ningún conjunto de datos."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c4fb7f20",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:12:20.821336Z",
     "iopub.status.busy": "2026-10-16T09:12:20.820927Z",
     "iopub.status.idle": "2026-10-16T09:12:24.471307Z",
     "shell.execute_reply": "2026-10-16T09:12:24.469883Z"
    }
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "import time, math, random\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.patches import Rectangle\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "torch.manual_seed(0); np.random.seed(0); random.seed(0)\n",
    "plt.rcParams[\"axes.grid\"] = False\n",
    "def count_params(m): return sum(p.numel() for p in m.parameters())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52e45e6c",
   "metadata": {},
   "source": [
    "## 1. Un mundo sintético de formas\n",
    "\n",
    "Cada imagen de 32×32 contiene de 1 a 3 formas (círculo, cuadrado o triángulo) de tamaño, posición y brillo aleatorios, sobre un fondo ruidoso e iluminado de forma desigual. Para cada imagen guardamos dos tipos de referencia:\n",
    "\n",
    "* una **máscara**: la clase de cada píxel (0 = fondo, 1 = círculo, 2 = cuadrado, 3 = triángulo);\n",
    "* una lista de **cajas**: $(x_{min}, y_{min}, x_{max}, y_{max}, \\text{clase})$ para cada objeto."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "650258e8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:12:24.475010Z",
     "iopub.status.busy": "2026-10-16T09:12:24.474579Z",
     "iopub.status.idle": "2026-10-16T09:12:26.226040Z",
     "shell.execute_reply": "2026-10-16T09:12:26.224275Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x520 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "proporción de píxeles por clase: {'fondo': '85.5%', 'círculo': '5.7%', 'cuadrado': '5.1%', 'triángulo': '3.7%'}\n"
     ]
    }
   ],
   "source": [
    "S = 32\n",
    "CLASSES = [\"fondo\", \"círculo\", \"cuadrado\", \"triángulo\"]\n",
    "yy, xx = np.mgrid[0:S, 0:S] + 0.5\n",
    "\n",
    "def shape_mask(kind, cx, cy, r):\n",
    "    if kind == 1:\n",
    "        return (xx - cx) ** 2 + (yy - cy) ** 2 <= r ** 2\n",
    "    if kind == 2:\n",
    "        return (np.abs(xx - cx) <= r * 0.85) & (np.abs(yy - cy) <= r * 0.85)\n",
    "    # triángulo hacia arriba inscrito en la caja [cx-r, cx+r] x [cy-r, cy+r]\n",
    "    top, base = cy - r, cy + r\n",
    "    half_width = r * (yy - top) / (2 * r)\n",
    "    return (yy >= top) & (yy <= base) & (np.abs(xx - cx) <= half_width)\n",
    "\n",
    "def make_scene(rng):\n",
    "    img = rng.uniform(0.0, 0.25) + 0.15 * (xx / S if rng.random() < 0.5 else yy / S)     # iluminación desigual\n",
    "    mask = np.zeros((S, S), dtype=np.int64)\n",
    "    boxes = []\n",
    "    for _ in range(rng.integers(1, 4)):\n",
    "        for attempt in range(20):                                   # se evitan solapamientos grandes\n",
    "            kind, r = int(rng.integers(1, 4)), rng.uniform(3.5, 7)\n",
    "            cx, cy = rng.uniform(r, S - r), rng.uniform(r, S - r)\n",
    "            m = shape_mask(kind, cx, cy, r)\n",
    "            if (mask[m] > 0).mean() < 0.1:\n",
    "                break\n",
    "        img = np.where(m, rng.uniform(0.55, 1.0), img)\n",
    "        mask[m] = kind\n",
    "        ys, xs = np.where(m)\n",
    "        boxes.append((xs.min(), ys.min(), xs.max() + 1, ys.max() + 1, kind))\n",
    "    img = img + rng.normal(0, 0.08, img.shape)\n",
    "    return img.astype(np.float32), mask, boxes\n",
    "\n",
    "def make_dataset(n, seed):\n",
    "    rng = np.random.default_rng(seed)\n",
    "    data = [make_scene(rng) for _ in range(n)]\n",
    "    X = torch.tensor(np.stack([d[0] for d in data])).unsqueeze(1)\n",
    "    M = torch.tensor(np.stack([d[1] for d in data]))\n",
    "    return X, M, [d[2] for d in data]\n",
    "\n",
    "X_tr, M_tr, B_tr = make_dataset(3000, 0)\n",
    "X_te, M_te, B_te = make_dataset(500, 1)\n",
    "colors = {1: \"tab:orange\", 2: \"tab:green\", 3: \"tab:purple\"}\n",
    "\n",
    "fig, axes = plt.subplots(2, 6, figsize=(15, 5.2))\n",
    "for k in range(6):\n",
    "    axes[0, k].imshow(X_tr[k, 0], cmap=\"gray\", vmin=0, vmax=1)\n",
    "    for x0, y0, x1, y1, c in B_tr[k]:\n",
    "        axes[0, k].add_patch(Rectangle((x0 - 0.5, y0 - 0.5), x1 - x0, y1 - y0, fill=False, color=colors[c], lw=2))\n",
    "    axes[1, k].imshow(M_tr[k], cmap=\"viridis\", vmin=0, vmax=3)\n",
    "    axes[0, k].axis(\"off\"); axes[1, k].axis(\"off\")\n",
    "axes[0, 0].set_title(\"imagen + cajas\", loc=\"left\"); axes[1, 0].set_title(\"máscara por píxel\", loc=\"left\")\n",
    "plt.show()\n",
    "print(\"proporción de píxeles por clase:\", {c: f\"{(M_tr == i).float().mean().item():.1%}\" for i, c in enumerate(CLASSES)})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6098e781",
   "metadata": {},
   "source": [
    "Fíjate en el desequilibrio entre clases: la mayoría de los píxeles son fondo. Un modelo que predijera \"fondo\" en todas partes ya tendría alrededor de un 80 % de \"exactitud\", y por eso la segmentación se evalúa por clase con la **IoU** (abajo), no con la exactitud por píxel.\n",
    "\n",
    "## 2. Segmentación semántica con una pequeña U-Net\n",
    "\n",
    "Un clasificador comprime la imagen hasta un solo vector. Un segmentador debe producir una salida **del mismo tamaño que la entrada**. La **U-Net** (Ronneberger et al., 2015) hace ambas cosas:\n",
    "\n",
    "* un **codificador** (camino de contracción) de convoluciones y pooling construye *qué* hay en la imagen, con una resolución cada vez menor;\n",
    "* un **decodificador** (camino de expansión) de convoluciones transpuestas vuelve a aumentar la resolución para decir *dónde*;\n",
    "* las **conexiones de salto** copian las características de alta resolución del codificador directamente al decodificador a la misma escala, para poder recuperar el detalle fino que el pooling perdió.\n",
    "\n",
    "```\n",
    "32×32 ─conv─► e1 (16c) ───────────────── salto ───────────────► concat ─conv─► 32×32 × 4 clases\n",
    "                │ pool                                             ▲ subir\n",
    "               16×16 ─conv─► e2 (32c) ─────── salto ───► concat ─conv─┘\n",
    "                               │ pool                      ▲ subir\n",
    "                              8×8 ─conv─► cuello de botella (64c) ┘\n",
    "```\n",
    "\n",
    "Una **convolución transpuesta** (`ConvTranspose2d`, paso 2) es el opuesto aprendible del pooling: cada píxel de entrada \"pinta\" un parche de 2×2 de la salida con pesos aprendidos."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d7c2ee00",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:12:26.229751Z",
     "iopub.status.busy": "2026-10-16T09:12:26.229475Z",
     "iopub.status.idle": "2026-10-16T09:16:14.783330Z",
     "shell.execute_reply": "2026-10-16T09:16:14.782215Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "U-Net                    parámetros 117,444  (125 s)\n",
      "   exactitud por píxel 99.9%   IoU fondo 1.00   IoU círculo 0.98   IoU cuadrado 0.98   IoU triángulo 0.98   IoU media (objetos) 0.98\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sin conexiones de salto  parámetros 105,924  (104 s)\n",
      "   exactitud por píxel 99.5%   IoU fondo 1.00   IoU círculo 0.95   IoU cuadrado 0.98   IoU triángulo 0.97   IoU media (objetos) 0.96\n"
     ]
    }
   ],
   "source": [
    "def double_conv(cin, cout):\n",
    "    return nn.Sequential(nn.Conv2d(cin, cout, 3, padding=1), nn.BatchNorm2d(cout), nn.ReLU(),\n",
    "                         nn.Conv2d(cout, cout, 3, padding=1), nn.BatchNorm2d(cout), nn.ReLU())\n",
    "\n",
    "class UNet(nn.Module):\n",
    "    def __init__(self, skips=True, n_classes=4):\n",
    "        super().__init__()\n",
    "        self.skips = skips\n",
    "        self.e1, self.e2, self.mid = double_conv(1, 16), double_conv(16, 32), double_conv(32, 64)\n",
    "        self.up2, self.up1 = nn.ConvTranspose2d(64, 32, 2, stride=2), nn.ConvTranspose2d(32, 16, 2, stride=2)\n",
    "        k = 2 if skips else 1\n",
    "        self.d2, self.d1 = double_conv(32 * k, 32), double_conv(16 * k, 16)\n",
    "        self.out = nn.Conv2d(16, n_classes, 1)                 # conv. 1x1: un clasificador en cada píxel\n",
    "    def forward(self, x):\n",
    "        e1 = self.e1(x)                                        # 32x32\n",
    "        e2 = self.e2(F.max_pool2d(e1, 2))                      # 16x16\n",
    "        m = self.mid(F.max_pool2d(e2, 2))                      # 8x8\n",
    "        u2 = self.up2(m)\n",
    "        d2 = self.d2(torch.cat([u2, e2], 1) if self.skips else u2)\n",
    "        u1 = self.up1(d2)\n",
    "        d1 = self.d1(torch.cat([u1, e1], 1) if self.skips else u1)\n",
    "        return self.out(d1)                                    # (lote, 4, 32, 32): puntuaciones de clase por píxel\n",
    "\n",
    "def iou_per_class(pred, mask, n=4):\n",
    "    return [((pred == c) & (mask == c)).sum().item() / max(1, ((pred == c) | (mask == c)).sum().item()) for c in range(n)]\n",
    "\n",
    "def fit_seg(model, epochs=8, lr=3e-3, batch=32):\n",
    "    opt = torch.optim.Adam(model.parameters(), lr=lr)\n",
    "    for ep in range(epochs):\n",
    "        model.train()\n",
    "        perm = torch.randperm(len(X_tr))\n",
    "        for i in range(0, len(X_tr), batch):\n",
    "            b = perm[i:i + batch]\n",
    "            loss = F.cross_entropy(model(X_tr[b]), M_tr[b])     # entropía cruzada promediada sobre todos los píxeles\n",
    "            opt.zero_grad(); loss.backward(); opt.step()\n",
    "    model.eval()\n",
    "    with torch.no_grad():\n",
    "        return model(X_te).argmax(1)\n",
    "\n",
    "seg = {}\n",
    "for name, skips in [(\"U-Net\", True), (\"sin conexiones de salto\", False)]:\n",
    "    torch.manual_seed(0)\n",
    "    t0 = time.time()\n",
    "    model = UNet(skips)\n",
    "    pred = fit_seg(model)\n",
    "    seg[name] = (model, pred)\n",
    "    ious = iou_per_class(pred, M_te)\n",
    "    print(f\"{name:24s} parámetros {count_params(model):,}  ({time.time() - t0:.0f} s)\\n   exactitud por píxel {(pred == M_te).float().mean().item():.1%}   \"\n",
    "          + \"   \".join(f\"IoU {c} {v:.2f}\" for c, v in zip(CLASSES, ious)) + f\"   IoU media (objetos) {np.mean(ious[1:]):.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f976901",
   "metadata": {},
   "source": [
    "La **intersección sobre la unión** (IoU) de una clase es $\\frac{|\\text{predicho} \\cap \\text{real}|}{|\\text{predicho} \\cup \\text{real}|}$: 1 es perfecto y, a diferencia de la exactitud por píxel, el enorme fondo no puede inflarla."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f14f7df6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:16:14.787487Z",
     "iopub.status.busy": "2026-10-16T09:16:14.787037Z",
     "iopub.status.idle": "2026-10-16T09:16:15.577242Z",
     "shell.execute_reply": "2026-10-16T09:16:15.575874Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x1000 with 24 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(4, 6, figsize=(15, 10))\n",
    "for k in range(6):\n",
    "    axes[0, k].imshow(X_te[k, 0], cmap=\"gray\", vmin=0, vmax=1)\n",
    "    axes[1, k].imshow(M_te[k], cmap=\"viridis\", vmin=0, vmax=3)\n",
    "    axes[2, k].imshow(seg[\"U-Net\"][1][k], cmap=\"viridis\", vmin=0, vmax=3)\n",
    "    axes[3, k].imshow(seg[\"sin conexiones de salto\"][1][k], cmap=\"viridis\", vmin=0, vmax=3)\n",
    "for ax, t in zip(axes[:, 0], [\"entrada\", \"máscara real\", \"U-Net\", \"sin saltos\"]):\n",
    "    ax.set_ylabel(t, fontsize=12)\n",
    "for ax in axes.ravel():\n",
    "    ax.set_xticks([]); ax.set_yticks([])\n",
    "plt.suptitle(\"morado azulado = fondo, verde azulado = círculo, verde = cuadrado, amarillo = triángulo\", y=0.92); plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "752f537d",
   "metadata": {},
   "source": [
    "Ambas redes funcionan bien con estas formas limpias; las conexiones de salto aportan un par de puntos de IoU, y la diferencia se concentra justo donde cabría esperar: en los bordes, las esquinas y las puntas de los triángulos, que el decodificador sin saltos tiene que reconstruir a partir de un resumen de 8×8. Con saltos, el decodificador sabe *qué* gracias al cuello de botella y *dónde* gracias a las características de alta resolución copiadas. En imágenes reales con estructura fina (vasos sanguíneos, bordes de carreteras) la diferencia es mucho mayor, y por eso la forma de U-Net se convirtió en estándar. La misma forma de codificador-decodificador con saltos se usa en imagen médica, en cartografía por satélite y como red de eliminación de ruido dentro de los modelos de difusión de imágenes.\n",
    "\n",
    "## 3. Detección de objetos al estilo YOLO\n",
    "\n",
    "La segmentación etiqueta píxeles, pero no separa dos objetos de la misma clase que se tocan ni los cuenta. La **detección** produce una caja por objeto.\n",
    "\n",
    "La idea de **YOLO** (\"You Only Look Once\", solo miras una vez; Redmon et al., 2015) convierte la detección en un único problema de regresión. Se divide la imagen en una cuadrícula de $S \\times S$ (aquí 4×4 celdas de 8×8 píxeles). La celda que contiene el **centro** de un objeto es la responsable de él, y predice:\n",
    "\n",
    "| Salida | Significado | Pérdida |\n",
    "|---|---|---|\n",
    "| $p_{obj}$ | ¿hay un objeto centrado en esta celda? | entropía cruzada binaria, todas las celdas |\n",
    "| $t_x, t_y$ | posición del centro dentro de la celda (0-1) | error cuadrático, solo celdas responsables |\n",
    "| $t_w, t_h$ | anchura y altura de la caja respecto a la imagen | error cuadrático sobre $\\sqrt{\\cdot}$, solo celdas responsables |\n",
    "| puntuaciones de clase | círculo / cuadrado / triángulo | entropía cruzada, solo celdas responsables |\n",
    "\n",
    "Así que toda la red es una CNN cuya salida es un tensor de $4 \\times 4 \\times 8$. (El YOLO real añade varias cajas \"ancla\" por celda para que una celda pueda contener más de un objeto; nosotros nos quedamos con una.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8871e7c1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:16:15.581633Z",
     "iopub.status.busy": "2026-10-16T09:16:15.581350Z",
     "iopub.status.idle": "2026-10-16T09:17:54.659288Z",
     "shell.execute_reply": "2026-10-16T09:17:54.657983Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "celdas con más de un centro de objeto (no se pueden representar, así que se ignoran): 31 de 6002 objetos de entrenamiento\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "detector: 109,688 parámetros, entrenado en 99 s, pérdida final 0.005\n"
     ]
    }
   ],
   "source": [
    "G, CELL = 4, S // 4\n",
    "\n",
    "def encode_targets(boxes_list):\n",
    "    T = torch.zeros(len(boxes_list), G, G, 6)                  # obj, tx, ty, tw, th, clase\n",
    "    for n, boxes in enumerate(boxes_list):\n",
    "        for x0, y0, x1, y1, c in boxes:\n",
    "            cx, cy = (x0 + x1) / 2, (y0 + y1) / 2\n",
    "            gx, gy = min(int(cx // CELL), G - 1), min(int(cy // CELL), G - 1)\n",
    "            T[n, gy, gx] = torch.tensor([1, cx / CELL - gx, cy / CELL - gy, (x1 - x0) / S, (y1 - y0) / S, c - 1])\n",
    "    return T\n",
    "\n",
    "T_tr, T_te = encode_targets(B_tr), encode_targets(B_te)\n",
    "print(f\"celdas con más de un centro de objeto (no se pueden representar, así que se ignoran): \"\n",
    "      f\"{sum(len(b) for b in B_tr) - int(T_tr[..., 0].sum())} de {sum(len(b) for b in B_tr)} objetos de entrenamiento\")\n",
    "\n",
    "class Detector(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.body = nn.Sequential(double_conv(1, 16), nn.MaxPool2d(2), double_conv(16, 32), nn.MaxPool2d(2),\n",
    "                                  double_conv(32, 64), nn.MaxPool2d(2), nn.Conv2d(64, 64, 3, padding=1), nn.ReLU())\n",
    "        self.head = nn.Conv2d(64, 1 + 4 + 3, 1)                # un vector de predicción por celda de la cuadrícula\n",
    "    def forward(self, x):\n",
    "        return self.head(self.body(x)).permute(0, 2, 3, 1)     # (lote, 4, 4, 8)\n",
    "\n",
    "def yolo_loss(out, T):\n",
    "    obj = T[..., 0] > 0\n",
    "    l_obj = F.binary_cross_entropy_with_logits(out[..., 0], T[..., 0])\n",
    "    xy = torch.sigmoid(out[..., 1:3][obj]); wh = torch.sigmoid(out[..., 3:5][obj])\n",
    "    l_box = F.mse_loss(xy, T[..., 1:3][obj]) + F.mse_loss(wh.sqrt(), T[..., 3:5][obj].sqrt())\n",
    "    l_cls = F.cross_entropy(out[..., 5:][obj], T[..., 5][obj].long())\n",
    "    return l_obj + 5 * l_box + l_cls\n",
    "\n",
    "torch.manual_seed(0)\n",
    "det = Detector()\n",
    "opt = torch.optim.Adam(det.parameters(), lr=2e-3)\n",
    "t0 = time.time()\n",
    "for ep in range(15):\n",
    "    det.train()\n",
    "    perm = torch.randperm(len(X_tr))\n",
    "    for i in range(0, len(X_tr), 32):\n",
    "        b = perm[i:i + 32]\n",
    "        loss = yolo_loss(det(X_tr[b]), T_tr[b])\n",
    "        opt.zero_grad(); loss.backward(); opt.step()\n",
    "print(f\"detector: {count_params(det):,} parámetros, entrenado en {time.time() - t0:.0f} s, pérdida final {loss.item():.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5bf1f61f",
   "metadata": {},
   "source": [
    "### De las salidas de la cuadrícula a las cajas: umbral y supresión de no máximos\n",
    "\n",
    "La decodificación convierte cada celda segura de nuevo en una caja en coordenadas de píxel. A veces dos celdas vecinas reclaman el mismo objeto, así que el último paso es la **supresión de no máximos (NMS)**: ordenar las cajas por confianza, quedarse con la mejor, borrar todas las cajas restantes de la misma clase que se solapen con ella más de un umbral de IoU, y repetir.\n",
    "\n",
    "Una caja predicha cuenta como **verdadero positivo** si se solapa con una caja de referencia aún no emparejada de la misma clase con IoU ≥ 0,5."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "16d7c816",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:17:54.662189Z",
     "iopub.status.busy": "2026-10-16T09:17:54.661892Z",
     "iopub.status.idle": "2026-10-16T09:17:57.702811Z",
     "shell.execute_reply": "2026-10-16T09:17:57.701622Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   umbral  precisión exhaust.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      0.9      98.8%    93.9%\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      0.7      98.9%    95.0%\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      0.5      98.9%    96.0%\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      0.3      98.9%    96.8%\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      0.1      98.9%    97.4%\n"
     ]
    }
   ],
   "source": [
    "def box_iou(a, b):\n",
    "    ix = max(0, min(a[2], b[2]) - max(a[0], b[0])); iy = max(0, min(a[3], b[3]) - max(a[1], b[1]))\n",
    "    inter = ix * iy\n",
    "    return inter / ((a[2] - a[0]) * (a[3] - a[1]) + (b[2] - b[0]) * (b[3] - b[1]) - inter + 1e-9)\n",
    "\n",
    "def decode(out, threshold=0.5, nms_iou=0.4):\n",
    "    boxes = []\n",
    "    p = torch.sigmoid(out[..., 0])\n",
    "    for gy in range(G):\n",
    "        for gx in range(G):\n",
    "            if p[gy, gx] > threshold:\n",
    "                tx, ty = torch.sigmoid(out[gy, gx, 1:3]).tolist(); tw, th = torch.sigmoid(out[gy, gx, 3:5]).tolist()\n",
    "                cx, cy, w, h = (gx + tx) * CELL, (gy + ty) * CELL, tw * S, th * S\n",
    "                boxes.append((cx - w / 2, cy - h / 2, cx + w / 2, cy + h / 2, int(out[gy, gx, 5:].argmax()) + 1, p[gy, gx].item()))\n",
    "    boxes.sort(key=lambda b: -b[5])\n",
    "    kept = []\n",
    "    for b in boxes:                                            # supresión de no máximos\n",
    "        if all(k[4] != b[4] or box_iou(k, b) < nms_iou for k in kept):\n",
    "            kept.append(b)\n",
    "    return kept\n",
    "\n",
    "def evaluate(threshold):\n",
    "    tp = fp = 0\n",
    "    n_true = sum(len(b) for b in B_te)\n",
    "    det.eval()\n",
    "    with torch.no_grad():\n",
    "        outs = det(X_te)\n",
    "    for out, truth in zip(outs, B_te):\n",
    "        unmatched = list(truth)\n",
    "        for b in decode(out, threshold):\n",
    "            match = next((t for t in unmatched if t[4] == b[4] and box_iou(t, b) >= 0.5), None)\n",
    "            if match: tp += 1; unmatched.remove(match)\n",
    "            else: fp += 1\n",
    "    return tp / max(1, tp + fp), tp / n_true\n",
    "\n",
    "print(f\"{'umbral':>9s} {'precisión':>10s} {'exhaust.':>8s}\")\n",
    "pr = []\n",
    "for th in [0.9, 0.7, 0.5, 0.3, 0.1]:\n",
    "    p, r = evaluate(th); pr.append((r, p))\n",
    "    print(f\"{th:9.1f} {p:10.1%} {r:8.1%}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7046c2e2",
   "metadata": {},
   "source": [
    "La **precisión** es la proporción de cajas predichas que son correctas; la **exhaustividad** (recall) es la proporción de objetos reales que se encontraron. El umbral de confianza intercambia una por otra, y los bancos de pruebas de detección resumen toda la curva como *precisión media promedio* (mAP). Veamos algunas detecciones:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f4f4a665",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:17:57.706904Z",
     "iopub.status.busy": "2026-10-16T09:17:57.706597Z",
     "iopub.status.idle": "2026-10-16T09:17:58.399638Z",
     "shell.execute_reply": "2026-10-16T09:17:58.398544Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x600 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "det.eval()\n",
    "with torch.no_grad():\n",
    "    outs = det(X_te[:12])\n",
    "fig, axes = plt.subplots(2, 6, figsize=(16, 6))\n",
    "for ax, img, out, truth in zip(axes.ravel(), X_te[:12], outs, B_te[:12]):\n",
    "    ax.imshow(img[0], cmap=\"gray\", vmin=0, vmax=1)\n",
    "    for x0, y0, x1, y1, c in truth:\n",
    "        ax.add_patch(Rectangle((x0 - 0.5, y0 - 0.5), x1 - x0, y1 - y0, fill=False, color=\"white\", lw=1, ls=\"--\"))\n",
    "    for x0, y0, x1, y1, c, conf in decode(out):\n",
    "        ax.add_patch(Rectangle((x0 - 0.5, y0 - 0.5), x1 - x0, y1 - y0, fill=False, color=colors[c], lw=2))\n",
    "        ax.text(x0, y0 - 1, f\"{CLASSES[c]} {conf:.2f}\", color=colors[c], fontsize=7)\n",
    "    for k in range(1, G):\n",
    "        ax.axhline(k * CELL - 0.5, color=\"gray\", lw=0.4); ax.axvline(k * CELL - 0.5, color=\"gray\", lw=0.4)\n",
    "    ax.set_xticks([]); ax.set_yticks([])\n",
    "plt.suptitle(\"Blanco discontinuo = referencia, en color = predicciones (las líneas grises muestran la cuadrícula de 4×4)\"); plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4d4e3f8",
   "metadata": {},
   "source": [
    "Una sola evaluación de la red por imagen da todas las cajas a la vez, y por eso los detectores al estilo YOLO funcionan en tiempo real. Su debilidad clásica se ve en el paso de codificación: una celda solo puede contener un objeto, así que los objetos pequeños muy juntos son difíciles. Los detectores modernos usan varias cajas ancla por celda, varias resoluciones de cuadrícula y (en los modelos tipo DETR) atención en lugar de una cuadrícula.\n",
    "\n",
    "## 4. Aprendizaje métrico: aprender qué significa \"lo mismo\"\n",
    "\n",
    "Un clasificador solo puede producir las clases con las que se entrenó. El reconocimiento facial no puede funcionar así: un teléfono debe reconocer a su dueño a partir de una sola foto de registro, sin volver a entrenar. La solución (FaceNet, Schroff et al., 2015; y antes las redes siamesas) es aprender un **embedding** $f(x)$ en el que las imágenes de la misma identidad estén cerca y las de identidades distintas, lejos. Entonces:\n",
    "\n",
    "* **verificación**: es la misma persona si $\\|f(x_1) - f(x_2)\\| < \\tau$;\n",
    "* **reconocimiento**: el embedding registrado más cercano.\n",
    "\n",
    "La **pérdida triplete** entrena esto directamente. Se toma un *ancla* $a$, un *positivo* $p$ (misma identidad) y un *negativo* $n$ (identidad distinta):\n",
    "\n",
    "$$ \\mathcal{L} = \\max\\big(0,\\; \\|f(a) - f(p)\\|^2 - \\|f(a) - f(n)\\|^2 + \\alpha\\big) $$\n",
    "\n",
    "Es cero en cuanto el negativo está más lejos que el positivo por al menos el margen $\\alpha$.\n",
    "\n",
    "Usamos dígitos manuscritos como \"identidades\" de sustitución. La red se entrena **solo con los dígitos del 0 al 6**. Después la probamos con imágenes nuevas de esos dígitos y con el **7, el 8 y el 9**, que nunca ha visto: la posición en la que está un sistema facial cuando conoce a una persona nueva."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7820bba7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:17:58.404044Z",
     "iopub.status.busy": "2026-10-16T09:17:58.403749Z",
     "iopub.status.idle": "2026-10-16T09:18:29.838917Z",
     "shell.execute_reply": "2026-10-16T09:18:29.837305Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "entrenado con los dígitos 0-6 durante 1500 lotes en 29 s, pérdida final 0.000\n"
     ]
    }
   ],
   "source": [
    "from sklearn.datasets import load_digits\n",
    "digits = load_digits()\n",
    "Xd = torch.tensor(digits.images / 16.0, dtype=torch.float32).unsqueeze(1)\n",
    "yd = torch.tensor(digits.target)\n",
    "seen, unseen = yd <= 6, yd >= 7\n",
    "g_split = torch.Generator().manual_seed(0)\n",
    "perm = torch.randperm(int(seen.sum()), generator=g_split)\n",
    "Xseen, yseen = Xd[seen][perm], yd[seen][perm]\n",
    "Xs, ys = Xseen[:900], yseen[:900]                 # imágenes de entrenamiento de los dígitos 0-6\n",
    "Xh, yh = Xseen[900:], yseen[900:]                 # imágenes reservadas de los MISMOS dígitos\n",
    "Xu, yu = Xd[unseen], yd[unseen]                   # dígitos 7-9, nunca vistos en el entrenamiento\n",
    "\n",
    "class Embedder(nn.Module):\n",
    "    def __init__(self, dim=16):\n",
    "        super().__init__()\n",
    "        self.net = nn.Sequential(nn.Conv2d(1, 32, 3, padding=1), nn.ReLU(), nn.Conv2d(32, 64, 3, padding=1), nn.ReLU(),\n",
    "                                 nn.MaxPool2d(2), nn.Flatten(), nn.Linear(64 * 16, 128), nn.ReLU(), nn.Linear(128, dim))\n",
    "    def forward(self, x):\n",
    "        return F.normalize(self.net(x), dim=1)                  # los embeddings viven en la esfera unidad\n",
    "\n",
    "def pk_batch(X, y, classes, P=7, K=8, gen=None):\n",
    "    # P identidades x K imágenes cada una, para que cada ancla tenga positivos y negativos en el mismo lote\n",
    "    idx = []\n",
    "    for c in classes[torch.randperm(len(classes), generator=gen)[:P]]:\n",
    "        pool = (y == c).nonzero().ravel()\n",
    "        idx.append(pool[torch.randint(0, len(pool), (K,), generator=gen)])\n",
    "    idx = torch.cat(idx)\n",
    "    return X[idx], y[idx]\n",
    "\n",
    "def batch_hard_triplet_loss(f, y, margin):\n",
    "    # para cada ancla: su positivo MÁS LEJANO y su negativo MÁS CERCANO del lote (el triplete más difícil)\n",
    "    D = torch.cdist(f, f).pow(2)\n",
    "    same = y[:, None] == y[None, :]\n",
    "    hardest_pos = (D * same).max(1).values\n",
    "    hardest_neg = (D + same * 1e9).min(1).values\n",
    "    return F.relu(hardest_pos - hardest_neg + margin).mean()\n",
    "\n",
    "torch.manual_seed(0)\n",
    "emb_net = Embedder()\n",
    "opt = torch.optim.Adam(emb_net.parameters(), lr=1e-3)\n",
    "gen = torch.Generator().manual_seed(0)\n",
    "classes, t0 = torch.arange(7), time.time()\n",
    "for step in range(1500):\n",
    "    xb, yb = pk_batch(Xs, ys, classes, gen=gen)\n",
    "    loss = batch_hard_triplet_loss(emb_net(xb), yb, margin=0.2)\n",
    "    opt.zero_grad(); loss.backward(); opt.step()\n",
    "print(f\"entrenado con los dígitos 0-6 durante 1500 lotes en {time.time() - t0:.0f} s, pérdida final {loss.item():.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1eced792",
   "metadata": {},
   "source": [
    "Los tripletes aleatorios son casi siempre *fáciles* (un \"1\" y un \"0\" ya están lejos), así que enseguida dejan de enseñar nada. La **minería de lo más difícil del lote** (batch-hard mining; Hermans et al., 2017) construye cada lote con unas pocas identidades con varias imágenes cada una y, para cada ancla, usa solo su positivo más difícil y su negativo más difícil dentro del lote.\n",
    "\n",
    "Ahora la prueba, en el escenario de un sistema de reconocimiento facial: **registrar una imagen por identidad** y después reconocer cada una de las demás imágenes por su embedding registrado más cercano. La ejecutamos dos veces: con imágenes nuevas de los dígitos vistos en el entrenamiento (0-6) y con los dígitos que la red nunca ha visto (7-9). Los píxeles en bruto son la referencia."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "42124ce2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-16T09:18:29.843061Z",
     "iopub.status.busy": "2026-10-16T09:18:29.842565Z",
     "iopub.status.idle": "2026-10-16T09:18:32.547637Z",
     "shell.execute_reply": "2026-10-16T09:18:32.546345Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x700 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "reconocimiento con UNA sola imagen de registro por identidad\n",
      "                        vistos 0-6 (7 clases)  no vistos 7-9 (3 clases)\n",
      "azar                                    14.3%                     33.3%\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "píxeles en bruto                        74.0%                     73.5%\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "embedding triplete                      99.2%                     68.2%\n"
     ]
    }
   ],
   "source": [
    "emb_net.eval()\n",
    "with torch.no_grad():\n",
    "    Eh, Eu = emb_net(Xh), emb_net(Xu)\n",
    "pix = lambda X: F.normalize(X.flatten(1), dim=1)\n",
    "\n",
    "def pair_distances(E, y):\n",
    "    D = torch.cdist(E, E)\n",
    "    same = (y[:, None] == y[None, :]) & ~torch.eye(len(y), dtype=torch.bool)\n",
    "    diff = y[:, None] != y[None, :]\n",
    "    return D[same].numpy(), D[diff].numpy()\n",
    "\n",
    "def one_shot(E, y, trials=200, seed=0):\n",
    "    # registra UN ejemplo aleatorio por identidad y clasifica cada uno de los demás por el embedding registrado más cercano\n",
    "    g = np.random.default_rng(seed); ids = sorted(set(y.tolist())); acc = []\n",
    "    for _ in range(trials):\n",
    "        refs = [g.choice(np.where(y.numpy() == d)[0]) for d in ids]\n",
    "        pred = torch.tensor(ids)[torch.cdist(E, E[refs]).argmin(1)]\n",
    "        mask = torch.ones(len(y), dtype=torch.bool); mask[refs] = False\n",
    "        acc.append((pred[mask] == y[mask]).float().mean().item())\n",
    "    return np.mean(acc)\n",
    "\n",
    "fig, axes = plt.subplots(2, 2, figsize=(15, 7))\n",
    "for row, (label, X, y, E) in enumerate([(\"dígitos vistos 0-6, imágenes nuevas\", Xh, yh, Eh), (\"dígitos NO VISTOS 7-9\", Xu, yu, Eu)]):\n",
    "    for col, (name, Z) in enumerate([(\"píxeles en bruto\", pix(X)), (\"embedding triplete\", E)]):\n",
    "        s, d = pair_distances(Z, y)\n",
    "        ax = axes[row, col]\n",
    "        ax.hist(s, bins=50, alpha=0.6, density=True, label=\"mismo dígito\"); ax.hist(d, bins=50, alpha=0.6, density=True, label=\"dígito distinto\")\n",
    "        ax.set_title(f\"{name}: {label}\", fontsize=10); ax.legend(fontsize=8)\n",
    "plt.tight_layout(); plt.show()\n",
    "\n",
    "print(\"reconocimiento con UNA sola imagen de registro por identidad\")\n",
    "print(f\"{'':22s} {'vistos 0-6 (7 clases)':>22s} {'no vistos 7-9 (3 clases)':>25s}\")\n",
    "print(f\"{'azar':22s} {1/7:22.1%} {1/3:25.1%}\")\n",
    "print(f\"{'píxeles en bruto':22s} {one_shot(pix(Xh), yh):22.1%} {one_shot(pix(Xu), yu):25.1%}\")\n",
    "print(f\"{'embedding triplete':22s} {one_shot(Eh, yh):22.1%} {one_shot(Eu, yu):25.1%}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "463cbf94",
   "metadata": {},
   "source": [
    "Con las identidades con las que se entrenó, el embedding es muchísimo mejor que los píxeles en bruto: las distancias \"iguales\" y \"distintas\" se separan limpiamente, y una imagen de registro basta. Con los dígitos no vistos el panorama es mucho menos halagador. Según la semilla aleatoria, el embedding queda en torno a la referencia de píxeles en bruto, a veces un poco por encima y a veces por debajo (prueba con varias semillas).\n",
    "\n",
    "Es un resultado honesto e importante, no un fallo. **Siete identidades de entrenamiento no bastan para aprender una noción general de \"lo mismo\".** Con tan pocas, la red aprende características que separan *esos siete dígitos* y no tiene ningún motivo para organizar bien el resto del espacio. FaceNet se entrenó con unos 8 millones de identidades precisamente por esto: generalizar a identidades nuevas es una propiedad que se compra con *diversidad de identidades*, no con más imágenes de las mismas pocas. La misma lección vale para cualquier modelo de embeddings, incluidos los embeddings de texto que hay detrás de la generación aumentada por recuperación (RAG): funcionan con entradas nuevas porque se entrenaron con una variedad enorme de ellas.\n",
    "\n",
    "## Ideas clave\n",
    "\n",
    "* **Segmentación** = un clasificador en cada píxel. Las redes codificador-decodificador con **conexiones de salto** (U-Net) recuperan el detalle que el pooling descarta. Se evalúa con la **IoU** por clase, no con la exactitud por píxel.\n",
    "* La **detección** puede plantearse como una sola regresión sobre una cuadrícula (YOLO): presencia de objeto + caja + clase por celda, decodificadas con un umbral y **supresión de no máximos**. La precisión y la exhaustividad se intercambian a través del umbral de confianza.\n",
    "* El **aprendizaje métrico** con una pérdida triplete aprende un embedding donde la distancia significa similitud, la base de la verificación facial y de la búsqueda por similitud. Generalizar a identidades nunca vistas en el entrenamiento requiere entrenar con *muchas* identidades.\n",
    "\n",
    "## Ejercicios\n",
    "\n",
    "1. Entrena la U-Net durante 2 épocas en lugar de 8. ¿Qué clase sufre más y por qué podrían ser los triángulos los más difíciles?\n",
    "2. Añade una cuarta forma (por ejemplo un anillo: un círculo con un agujero) a `shape_mask` y vuelve a entrenar el segmentador y el detector.\n",
    "3. En el detector, cambia la cuadrícula a 8×8 celdas de 4 píxeles. ¿Cómo cambian la precisión y la exhaustividad, y por qué las cuadrículas finas ayudan en escenas abarrotadas?\n",
    "4. Quita la NMS de `decode` (conserva todas las cajas por encima del umbral). ¿Cómo cambia la precisión?\n",
    "5. Entrena la red triplete con márgenes de 0,1 y 0,5, y con tres semillas aleatorias para cada uno. ¿Qué tan estable es la exactitud con un solo ejemplo en los dígitos no vistos, comparada con los vistos?\n",
    "6. Entrena solo con los dígitos 0-4 y prueba con 5-9, y después entrena con 0-7 y prueba con 8-9. ¿Mejora la exactitud con identidades no vistas a medida que crece el número de identidades de entrenamiento?"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
