Spaces:
Running
Running
File size: 5,180 Bytes
22e1b62 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 |
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"import polars as pl\n",
"import matplotlib.pyplot as plt\n",
"import seaborn as sns\n",
"from sklearn.metrics import roc_auc_score, accuracy_score, precision_score, recall_score, f1_score, RocCurveDisplay\n",
"\n",
"sns.set()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def pfbeta(labels, predictions, beta=1):\n",
" y_true_count = 0\n",
" ctp = 0\n",
" cfp = 0\n",
"\n",
" for idx in range(len(labels)):\n",
" prediction = min(max(predictions[idx], 0), 1)\n",
" if (labels[idx]):\n",
" y_true_count += 1\n",
" ctp += prediction\n",
" else:\n",
" cfp += prediction\n",
"\n",
" beta_squared = beta * beta\n",
" c_precision = ctp / (ctp + cfp)\n",
" c_recall = ctp / y_true_count\n",
" if (c_precision > 0 and c_recall > 0):\n",
" result = (1 + beta_squared) * (c_precision * c_recall) / (beta_squared * c_precision + c_recall)\n",
" return result\n",
" else:\n",
" return 0"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def get_part_metrics(df: pl.DataFrame, threshold=0.3) -> dict:\n",
" df = df.with_columns((df[\"preds\"] > threshold).alias(\"preds_bin\"))\n",
" metrics = {}\n",
" # binary metrics using the threshold\n",
" metrics[\"accuracy\"] = accuracy_score(df[\"labels\"].to_numpy(), df[\"preds_bin\"].to_numpy())\n",
" metrics[\"precision\"] = precision_score(df[\"labels\"].to_numpy(), df[\"preds_bin\"].to_numpy())\n",
" metrics[\"recall\"] = recall_score(df[\"labels\"].to_numpy(), df[\"preds_bin\"].to_numpy())\n",
" metrics[\"f1\"] = f1_score(df[\"labels\"].to_numpy(), df[\"preds_bin\"].to_numpy())\n",
" # probabilistic F1 (doesn't depend on the threshold)\n",
" metrics[\"pf1\"] = pfbeta(df[\"labels\"].to_numpy(), df[\"preds\"].to_numpy())\n",
" # ROC AUC\n",
" metrics[\"roc_auc\"] = roc_auc_score(df[\"labels\"].to_numpy(), df[\"preds\"].to_numpy())\n",
" return metrics\n",
"\n",
"\n",
"def get_all_metrics(df: pl.DataFrame, threshold=0.3) -> pd.DataFrame:\n",
" groups = [list(range(5)), [0, 1], [0, 4], [0, 2], [0, 3]]\n",
" group_names = [\"all\", \"StableDiffusion\", \"Midjourney\", \"Dalle2\", \"Dalle3\"]\n",
" all_metrics = []\n",
" for i, g in enumerate(groups):\n",
" subset = df.filter(pl.col(\"domains\").is_in(g))\n",
" metrics = get_part_metrics(subset, threshold=threshold)\n",
" metrics[\"group\"] = group_names[i]\n",
" all_metrics.append(metrics)\n",
" \n",
" return pd.DataFrame(all_metrics)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df1 = pl.read_csv(\"outputs/preds-image-classifier-1.csv\")\n",
"metrics_df1 = get_all_metrics(df1, threshold=0.5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metrics_df1"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df14 = pl.read_csv(\"outputs/preds-image-classifier-14.csv\")\n",
"metrics_df14 = get_all_metrics(df14, threshold=0.5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metrics_df14"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df142 = pl.read_csv(\"outputs/preds-image-classifier-142.csv\")\n",
"metrics_df142 = get_all_metrics(df142, threshold=0.5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metrics_df142"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df1423 = pl.read_csv(\"outputs/preds-image-classifier-1423.csv\")\n",
"metrics_df1423 = get_all_metrics(df1423, threshold=0.5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metrics_df1423"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "GenAI-image-detection-Z_9oJJe7",
"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": "3.11.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|