mirror of
https://github.com/nerfstudio-project/nerfstudio.git
synced 2026-11-11 02:24:38 +08:00
196 lines
6.7 KiB
Plaintext
196 lines
6.7 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"%load_ext autoreload\n",
|
|
"%autoreload 2"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 3,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"import pandas as pd\n",
|
|
"import copy\n",
|
|
"from benchmark_nerfacto import table_rows\n",
|
|
"\n",
|
|
"# methods sorted in the same order we want in the paper tables\n",
|
|
"methods = [row[0] for row in table_rows]"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 4,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def csv_filename_to_results(csv_filename, num_decimals=2):\n",
|
|
" # code to read the csv file\n",
|
|
" # and return a dictionary\n",
|
|
"\n",
|
|
" df = pd.read_csv(csv_filename)\n",
|
|
"\n",
|
|
" column_titles = list(df.columns.values)\n",
|
|
" last_row = list(df.values[-1])\n",
|
|
"\n",
|
|
" # print(len(column_titles))\n",
|
|
" # print(len(last_row))\n",
|
|
"\n",
|
|
" from collections import defaultdict\n",
|
|
" results = defaultdict(dict)\n",
|
|
" for column_title, last_row_value in zip(column_titles, last_row):\n",
|
|
" if column_title == \"Step\" or \"__MAX\" in column_title or \"__MIN\" in column_title:\n",
|
|
" continue\n",
|
|
" name = column_title.split(\" - \")[0]\n",
|
|
" capture_name, method = name.split(\"_\")\n",
|
|
" metric = float(last_row_value)\n",
|
|
" # format to have 2 decimal places\n",
|
|
" if num_decimals == 2:\n",
|
|
" metric = f\"{metric:.2f}\"\n",
|
|
" elif num_decimals == 3:\n",
|
|
" metric = f\"{metric:.3f}\"\n",
|
|
" results[capture_name][method] = metric\n",
|
|
"\n",
|
|
" # import pprint\n",
|
|
" # pprint.pprint(results)\n",
|
|
" # print(len(results))\n",
|
|
"\n",
|
|
" results = dict(results)\n",
|
|
" return results"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 11,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# the csv filenames\n",
|
|
"# these are generated from exporting CSVs from wandb graphs\n",
|
|
"# csv_filename_psrn = \"/home/ethanweber/nerfactory/wandb-psnr.csv\"\n",
|
|
"# csv_filename_ssim = \"/home/ethanweber/nerfactory/wandb-ssim.csv\"\n",
|
|
"# csv_filename_lpips = \"/home/ethanweber/nerfactory/wandb-lpips.csv\"\n",
|
|
"csv_filename_psrn = \"/home/ethanweber/nerfactory/wandb-jan9-psnr.csv\"\n",
|
|
"csv_filename_ssim = \"/home/ethanweber/nerfactory/wandb-jan9-ssim.csv\"\n",
|
|
"csv_filename_lpips = \"/home/ethanweber/nerfactory/wandb-jan9-lpips.csv\"\n",
|
|
"\n",
|
|
"results_psrn = csv_filename_to_results(csv_filename_psrn)\n",
|
|
"results_ssim = csv_filename_to_results(csv_filename_ssim, num_decimals=3)\n",
|
|
"results_lpips = csv_filename_to_results(csv_filename_lpips, num_decimals=3)\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 12,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def generate_row(method, capture_names):\n",
|
|
" # remove spaces from method\n",
|
|
" method_text = method.replace(\"-\", \" \")\n",
|
|
" string = f\"{method_text}\"\n",
|
|
" for idx, capture_name in enumerate(capture_names):\n",
|
|
" psnr = results_psrn[capture_name][method]\n",
|
|
" ssim = results_ssim[capture_name][method]\n",
|
|
" lpips = results_lpips[capture_name][method]\n",
|
|
" if idx == len(capture_names) - 1:\n",
|
|
" string += \" & \\\\multicolumn{1}{r|}{\" + f\"{psnr} / {ssim} / {lpips}\" + \"}\"\n",
|
|
" else:\n",
|
|
" string += \" & \\\\multicolumn{1}{r}{\" + f\"{psnr} / {ssim} / {lpips}\" + \"}\"\n",
|
|
" return string"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 13,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"['Egypt', 'person', 'kitchen', 'plane', 'dozer']\n",
|
|
"no contraction & \\multicolumn{1}{r}{18.06 / 0.520 / 0.151} & \\multicolumn{1}{r}{21.96 / 0.511 / 0.125} & \\multicolumn{1}{r}{19.35 / 0.764 / 0.078} & \\multicolumn{1}{r}{18.45 / 0.519 / 0.161} & \\multicolumn{1}{r|}{19.53 / 0.786 / 0.081} \\\\ \n",
|
|
"synthetic on real & \\multicolumn{1}{r}{18.51 / 0.546 / 0.119} & \\multicolumn{1}{r}{24.68 / 0.667 / 0.099} & \\multicolumn{1}{r}{20.98 / 0.745 / 0.074} & \\multicolumn{1}{r}{19.32 / 0.578 / 0.102} & \\multicolumn{1}{r|}{21.41 / 0.758 / 0.080} \\\\ \\hline\n",
|
|
"\n",
|
|
"['floating-tree', 'aspen', 'stump', 'sculpture', 'Giannini-Hall']\n",
|
|
"no contraction & \\multicolumn{1}{r}{18.98 / 0.679 / 0.088} & \\multicolumn{1}{r}{15.80 / 0.330 / 0.342} & \\multicolumn{1}{r}{20.69 / 0.704 / 0.075} & \\multicolumn{1}{r}{21.43 / 0.720 / 0.073} & \\multicolumn{1}{r|}{17.75 / 0.596 / 0.134} \\\\ \n",
|
|
"synthetic on real & \\multicolumn{1}{r}{21.13 / 0.726 / 0.072} & \\multicolumn{1}{r}{15.53 / 0.301 / 0.332} & \\multicolumn{1}{r}{23.86 / 0.786 / 0.053} & \\multicolumn{1}{r}{22.62 / 0.706 / 0.076} & \\multicolumn{1}{r|}{18.41 / 0.576 / 0.133} \\\\ \\hline\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"capture_names_row1 = [\n",
|
|
" \"Egypt\",\n",
|
|
" \"person\",\n",
|
|
" \"kitchen\",\n",
|
|
" \"plane\",\n",
|
|
" \"dozer\"\n",
|
|
"]\n",
|
|
"capture_names_row2 = [\n",
|
|
" \"floating-tree\",\n",
|
|
" \"aspen\",\n",
|
|
" \"stump\",\n",
|
|
" \"sculpture\",\n",
|
|
" \"Giannini-Hall\"\n",
|
|
"]\n",
|
|
"\n",
|
|
"print(capture_names_row1)\n",
|
|
"for method in methods[:-1]:\n",
|
|
" # print(generate_row(method, capture_names_row1) + ' \\\\\\ \\cline{1-1}')\n",
|
|
" print(generate_row(method, capture_names_row1) + ' \\\\\\ ')\n",
|
|
"print(generate_row(methods[-1], capture_names_row1) + ' \\\\\\ \\hline')\n",
|
|
"\n",
|
|
"print()\n",
|
|
"\n",
|
|
"print(capture_names_row2)\n",
|
|
"for method in methods[:-1]:\n",
|
|
" # print(generate_row(method, capture_names_row2) + ' \\\\\\ \\cline{1-1}')\n",
|
|
" print(generate_row(method, capture_names_row2) + ' \\\\\\ ')\n",
|
|
"print(generate_row(methods[-1], capture_names_row2) + ' \\\\\\ \\hline')\n",
|
|
"\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": []
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "nerfactory",
|
|
"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": "3.10.6"
|
|
},
|
|
"orig_nbformat": 4,
|
|
"vscode": {
|
|
"interpreter": {
|
|
"hash": "b7b86c17c0dfd6b313529972d424f2e075cc198a46cda005757126afe8133c7b"
|
|
}
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 2
|
|
}
|