From 4c100299e961efdf599ea282280998f4a5c15860 Mon Sep 17 00:00:00 2001 From: xfr Date: Mon, 29 Jul 2019 14:33:09 +0800 Subject: [PATCH] add libra ga --- PrepareData.ipynb | 18 +- analyse.ipynb | 1423 +++++++++++---------- configs/rscup/ga.py | 174 +++ configs/rscup/htc_anchor.py | 307 +++++ configs/rscup/htc_deform_focal.py | 4 +- configs/rscup/htc_libra.py | 313 +++++ configs/rscup/htc_sy.py | 307 +++++ generate_chips.ipynb | 42 +- mmdet/apis/inference.py | 14 +- mmdet/models/anchor_heads/ga_rpn_head.py | 1 + mmdet/models/losses/cross_entropy_loss.py | 10 +- pipeline.ipynb | 204 ++- tools/demo.py | 13 +- 13 files changed, 1972 insertions(+), 858 deletions(-) create mode 100644 configs/rscup/ga.py create mode 100644 configs/rscup/htc_anchor.py create mode 100644 configs/rscup/htc_libra.py create mode 100644 configs/rscup/htc_sy.py diff --git a/PrepareData.ipynb b/PrepareData.ipynb index 7782c99..fa27245 100644 --- a/PrepareData.ipynb +++ b/PrepareData.ipynb @@ -2,7 +2,7 @@ "cells": [ { "cell_type": "code", - "execution_count": 119, + "execution_count": 3, "metadata": {}, "outputs": [ { @@ -798,7 +798,7 @@ }, { "cell_type": "code", - "execution_count": 166, + "execution_count": 1, "metadata": {}, "outputs": [ { @@ -806,7 +806,7 @@ "output_type": "stream", "text": [ "loading annotations into memory...\n", - "Done (t=6.87s)\n", + "Done (t=6.40s)\n", "creating index...\n", "index created!\n" ] @@ -824,7 +824,7 @@ }, { "cell_type": "code", - "execution_count": 182, + "execution_count": 7, "metadata": {}, "outputs": [ { @@ -832,23 +832,23 @@ "output_type": "stream", "text": [ "[16]\n", - "{'license': 1, 'file_name': 'P4724_1_3328.0_2496.0_0part30.jpg', 'coco_url': 'xxx', 'height': 511.0, 'width': 511.0, 'date_captured': '2019-06-25', 'flickr_url': 'xxx', 'id': 13547}\n", - "2\n" + "{'license': 1, 'file_name': 'P1933_0.5_292.0_191.0_0part3.jpg', 'coco_url': 'xxx', 'height': 511.0, 'width': 511.0, 'date_captured': '2019-06-25', 'flickr_url': 'xxx', 'id': 42637}\n", + "132\n" ] }, { "data": { "text/plain": [ - "" + "" ] }, - "execution_count": 182, + "execution_count": 7, "metadata": {}, "output_type": "execute_result" }, { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] diff --git a/analyse.ipynb b/analyse.ipynb index ea97a2a..81f017a 100644 --- a/analyse.ipynb +++ b/analyse.ipynb @@ -2,7 +2,7 @@ "cells": [ { "cell_type": "code", - "execution_count": 2, + "execution_count": 1, "metadata": { "collapsed": true }, @@ -3153,7 +3153,7 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 9, "metadata": {}, "outputs": [ { @@ -3161,15 +3161,16 @@ "output_type": "stream", "text": [ "loading annotations into memory...\n", - "Done (t=2.65s)\n", + "Done (t=5.74s)\n", "creating index...\n", - "index created!\n" + "index created!\n", + "23026\n" ] }, { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "fe97a10be21c42d0ac0bbc990bbd7421", + "model_id": "8581c993987c4de2a4a684fd3fb1d6bd", "version_major": 2, "version_minor": 0 }, @@ -3185,8 +3186,8 @@ "output_type": "stream", "text": [ "\n", - "{'tennis-court': 2777, 'container-crane': 256, 'storage-tank': 7442, 'baseball-diamond': 692, 'plane': 8533, 'ground-track-field': 498, 'helicopter': 652, 'airport': 306, 'harbor': 6459, 'ship': 40552, 'large-vehicle': 24570, 'swimming-pool': 2379, 'soccer-ball-field': 421, 'roundabout': 657, 'basketball-court': 564, 'bridge': 2497, 'small-vehicle': 169268, 'helipad': 104}\n", - "{'license': 1, 'file_name': 'P7763.png', 'coco_url': 'xxx', 'height': 4096, 'width': 4096, 'date_captured': '2019-06-25', 'flickr_url': 'xxx', 'id': 1559}\n" + "{'tennis-court': 7505, 'container-crane': 35, 'storage-tank': 2799, 'baseball-diamond': 527, 'plane': 8851, 'ground-track-field': 68, 'helicopter': 479, 'airport': 0, 'harbor': 13382, 'ship': 88896, 'large-vehicle': 86480, 'swimming-pool': 6956, 'soccer-ball-field': 270, 'roundabout': 336, 'basketball-court': 858, 'bridge': 137, 'small-vehicle': 336870, 'helipad': 7}\n", + "{'license': 1, 'file_name': 'P7885_1_1248.0_0.0_0part21.jpg', 'coco_url': 'xxx', 'height': 511.0, 'width': 511.0, 'date_captured': '2019-06-25', 'flickr_url': 'xxx', 'id': 29705}\n" ] } ], @@ -3194,10 +3195,12 @@ "import matplotlib.pyplot as plt\n", "CLASS=['tennis-court', 'container-crane', 'storage-tank', 'baseball-diamond', 'plane', 'ground-track-field', 'helicopter', 'airport', 'harbor', 'ship', 'large-vehicle', 'swimming-pool', 'soccer-ball-field', 'roundabout', 'basketball-court', 'bridge', 'small-vehicle', 'helipad']\n", "#CLASS={'tennis-court', 'container-crane', 'storage-tank', 'baseball-diamond', 'plane', 'ground-track-field', 'helicopter', 'airport', 'harbor', 'ship', 'large-vehicle', 'swimming-pool', 'soccer-ball-field', 'roundabout', 'basketball-court', 'bridge', 'small-vehicle', 'helipad'}\n", - "coco=COCO(\"/home/xfr/rssid/data/annotation/annos_rscup_train.json\")\n", + "coco=COCO(\"/home/xfr/rssid/rscup/annotation/annos_rscup_train.json\")\n", "class_to_ind = dict(zip(CLASS, range(len(CLASS))))\n", "num_class = dict(zip(CLASS, [0]*len(CLASS)))\n", - "imgIds = coco.getImgIds()\n", + "catIds = coco.getCatIds(catNms=['small-vehicle'])\n", + "imgIds = coco.getImgIds(catIds=catIds)\n", + "print(len(imgIds))\n", "max_num = []\n", "for cls in tqdm(CLASS):\n", " \n", @@ -3217,152 +3220,349 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "[1017,\n", - " 869,\n", - " 741,\n", - " 631,\n", - " 626,\n", - " 595,\n", - " 586,\n", - " 586,\n", - " 515,\n", - " 511,\n", - " 510,\n", - " 509,\n", - " 504,\n", - " 497,\n", - " 497,\n", - " 463,\n", - " 429,\n", - " 429,\n", - " 422,\n", - " 412,\n", - " 408,\n", + "[1162,\n", + " 719,\n", + " 634,\n", + " 622,\n", + " 561,\n", + " 503,\n", + " 502,\n", + " 496,\n", + " 473,\n", + " 470,\n", + " 468,\n", + " 465,\n", + " 465,\n", + " 465,\n", + " 462,\n", + " 459,\n", + " 449,\n", + " 446,\n", + " 440,\n", + " 438,\n", + " 435,\n", + " 434,\n", + " 430,\n", + " 424,\n", + " 419,\n", + " 415,\n", + " 406,\n", " 404,\n", - " 395,\n", - " 394,\n", " 388,\n", - " 386,\n", - " 384,\n", + " 387,\n", " 379,\n", - " 378,\n", - " 348,\n", + " 373,\n", + " 371,\n", + " 371,\n", + " 366,\n", + " 360,\n", + " 356,\n", + " 356,\n", + " 354,\n", + " 354,\n", + " 353,\n", + " 353,\n", + " 347,\n", + " 347,\n", + " 345,\n", + " 345,\n", " 343,\n", - " 340,\n", - " 340,\n", + " 342,\n", + " 341,\n", + " 341,\n", + " 338,\n", + " 334,\n", + " 333,\n", + " 332,\n", + " 331,\n", " 330,\n", - " 325,\n", - " 325,\n", - " 325,\n", + " 330,\n", + " 324,\n", + " 324,\n", + " 322,\n", + " 319,\n", + " 318,\n", " 312,\n", " 310,\n", + " 310,\n", + " 309,\n", " 306,\n", - " 301,\n", + " 305,\n", + " 304,\n", + " 299,\n", + " 299,\n", + " 298,\n", + " 296,\n", " 296,\n", " 295,\n", + " 294,\n", + " 293,\n", + " 292,\n", + " 290,\n", + " 289,\n", + " 288,\n", + " 288,\n", + " 287,\n", " 287,\n", + " 286,\n", + " 286,\n", " 285,\n", + " 284,\n", + " 284,\n", + " 284,\n", + " 283,\n", + " 282,\n", " 281,\n", + " 279,\n", + " 277,\n", " 276,\n", " 274,\n", - " 273,\n", + " 272,\n", " 272,\n", " 269,\n", + " 268,\n", + " 268,\n", + " 268,\n", + " 268,\n", + " 267,\n", " 267,\n", - " 265,\n", + " 266,\n", + " 264,\n", + " 264,\n", + " 263,\n", + " 263,\n", + " 263,\n", " 262,\n", - " 258,\n", + " 256,\n", + " 256,\n", + " 255,\n", + " 255,\n", + " 255,\n", + " 254,\n", + " 253,\n", + " 252,\n", + " 252,\n", " 251,\n", - " 250,\n", " 249,\n", - " 248,\n", - " 243,\n", + " 249,\n", + " 247,\n", + " 246,\n", + " 246,\n", + " 246,\n", + " 246,\n", + " 242,\n", + " 242,\n", + " 242,\n", + " 242,\n", + " 242,\n", + " 242,\n", + " 241,\n", + " 241,\n", + " 240,\n", + " 236,\n", " 236,\n", " 235,\n", + " 235,\n", + " 234,\n", + " 233,\n", + " 233,\n", + " 233,\n", + " 233,\n", + " 233,\n", + " 233,\n", " 232,\n", " 232,\n", + " 231,\n", " 230,\n", - " 225,\n", + " 230,\n", + " 230,\n", + " 229,\n", + " 228,\n", + " 228,\n", + " 227,\n", + " 226,\n", + " 226,\n", " 225,\n", " 224,\n", + " 224,\n", + " 223,\n", " 223,\n", + " 223,\n", + " 222,\n", + " 222,\n", + " 222,\n", + " 222,\n", + " 222,\n", + " 222,\n", + " 221,\n", " 221,\n", - " 220,\n", + " 221,\n", + " 219,\n", " 219,\n", " 218,\n", + " 218,\n", " 217,\n", " 215,\n", + " 215,\n", + " 215,\n", + " 214,\n", " 214,\n", + " 214,\n", + " 214,\n", + " 213,\n", + " 213,\n", " 212,\n", " 212,\n", " 212,\n", - " 209,\n", + " 212,\n", + " 212,\n", + " 212,\n", + " 210,\n", " 209,\n", " 207,\n", + " 207,\n", + " 206,\n", + " 206,\n", + " 206,\n", + " 206,\n", + " 205,\n", + " 205,\n", + " 205,\n", " 204,\n", - " 202,\n", - " 202,\n", - " 199,\n", - " 199,\n", - " 199,\n", - " 199,\n", - " 197,\n", - " 197,\n", + " 204,\n", + " 200,\n", + " 200,\n", " 197,\n", " 197,\n", " 196,\n", - " 195,\n", + " 196,\n", " 194,\n", " 193,\n", + " 193,\n", + " 193,\n", + " 192,\n", + " 192,\n", " 191,\n", " 190,\n", + " 190,\n", + " 190,\n", + " 190,\n", + " 190,\n", + " 189,\n", + " 189,\n", + " 188,\n", " 188,\n", " 188,\n", + " 188,\n", + " 187,\n", " 187,\n", + " 187,\n", + " 187,\n", + " 186,\n", + " 186,\n", + " 186,\n", + " 186,\n", + " 186,\n", + " 186,\n", " 185,\n", " 185,\n", + " 185,\n", + " 184,\n", " 184,\n", + " 184,\n", + " 184,\n", + " 182,\n", + " 182,\n", " 182,\n", + " 180,\n", + " 180,\n", + " 179,\n", + " 179,\n", + " 179,\n", + " 179,\n", + " 179,\n", + " 179,\n", " 178,\n", + " 178,\n", + " 177,\n", + " 177,\n", + " 177,\n", " 177,\n", " 176,\n", " 175,\n", " 175,\n", - " 175,\n", + " 174,\n", + " 174,\n", + " 173,\n", + " 173,\n", + " 173,\n", " 173,\n", + " 172,\n", + " 172,\n", " 171,\n", + " 171,\n", + " 171,\n", + " 171,\n", + " 170,\n", " 170,\n", + " 169,\n", " 168,\n", " 167,\n", - " 167,\n", + " 166,\n", + " 166,\n", + " 166,\n", + " 166,\n", + " 166,\n", " 166,\n", " 166,\n", " 166,\n", " 165,\n", " 165,\n", - " 165,\n", - " 165,\n", + " 164,\n", + " 164,\n", + " 164,\n", " 163,\n", " 163,\n", + " 163,\n", + " 162,\n", + " 162,\n", + " 162,\n", + " 162,\n", " 162,\n", " 162,\n", " 161,\n", - " 160,\n", - " 160,\n", - " 158,\n", + " 161,\n", + " 161,\n", + " 159,\n", + " 159,\n", + " 159,\n", + " 159,\n", + " 159,\n", + " 159,\n", " 158,\n", " 158,\n", + " 157,\n", + " 157,\n", + " 157,\n", + " 157,\n", + " 157,\n", + " 156,\n", + " 156,\n", + " 156,\n", + " 155,\n", + " 155,\n", " 155,\n", " 155,\n", " 155,\n", " 154,\n", - " 153,\n", + " 154,\n", + " 154,\n", " 153,\n", " 153,\n", " 152,\n", @@ -3371,67 +3571,201 @@ " 151,\n", " 150,\n", " 150,\n", + " 150,\n", + " 149,\n", + " 149,\n", " 149,\n", " 149,\n", " 149,\n", + " 149,\n", + " 149,\n", + " 148,\n", + " 148,\n", " 148,\n", " 147,\n", + " 147,\n", + " 147,\n", + " 147,\n", + " 147,\n", + " 146,\n", + " 146,\n", + " 146,\n", " 146,\n", - " 145,\n", - " 145,\n", - " 145,\n", " 145,\n", " 144,\n", " 144,\n", - " 144,\n", - " 143,\n", " 143,\n", " 143,\n", " 142,\n", + " 142,\n", + " 142,\n", + " 142,\n", + " 142,\n", + " 142,\n", " 141,\n", " 141,\n", " 140,\n", " 140,\n", + " 140,\n", + " 140,\n", + " 140,\n", + " 140,\n", + " 140,\n", + " 140,\n", + " 139,\n", " 139,\n", " 139,\n", + " 139,\n", + " 139,\n", + " 139,\n", + " 139,\n", + " 139,\n", + " 139,\n", + " 139,\n", + " 138,\n", " 138,\n", + " 138,\n", + " 138,\n", + " 138,\n", + " 138,\n", + " 137,\n", + " 137,\n", + " 137,\n", + " 137,\n", " 137,\n", + " 137,\n", + " 137,\n", + " 137,\n", + " 137,\n", + " 137,\n", + " 137,\n", + " 136,\n", + " 136,\n", + " 136,\n", + " 136,\n", + " 136,\n", + " 135,\n", + " 135,\n", " 135,\n", " 135,\n", + " 135,\n", + " 135,\n", + " 134,\n", + " 134,\n", " 134,\n", " 134,\n", " 134,\n", " 133,\n", " 133,\n", + " 133,\n", + " 133,\n", + " 133,\n", + " 133,\n", + " 133,\n", + " 133,\n", + " 132,\n", + " 132,\n", + " 132,\n", " 132,\n", " 131,\n", " 131,\n", " 131,\n", - " 131,\n", - " 131,\n", " 130,\n", + " 130,\n", + " 130,\n", + " 130,\n", + " 129,\n", + " 129,\n", + " 129,\n", + " 129,\n", + " 129,\n", + " 129,\n", " 129,\n", " 129,\n", + " 129,\n", + " 128,\n", + " 128,\n", + " 128,\n", + " 128,\n", + " 128,\n", + " 128,\n", + " 128,\n", " 128,\n", " 127,\n", " 127,\n", " 127,\n", " 127,\n", + " 127,\n", + " 127,\n", + " 127,\n", + " 127,\n", + " 127,\n", + " 127,\n", + " 126,\n", + " 126,\n", " 126,\n", + " 126,\n", + " 125,\n", + " 125,\n", + " 125,\n", " 125,\n", + " 125,\n", + " 125,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", + " 124,\n", " 124,\n", " 124,\n", " 123,\n", + " 123,\n", + " 123,\n", + " 122,\n", + " 122,\n", + " 122,\n", + " 122,\n", " 122,\n", " 122,\n", " 121,\n", + " 121,\n", + " 121,\n", + " 121,\n", + " 121,\n", + " 121,\n", + " 121,\n", + " 121,\n", " 120,\n", " 120,\n", " 120,\n", " 120,\n", - " 120,\n", - " 120,\n", " 119,\n", + " 119,\n", + " 119,\n", + " 119,\n", + " 119,\n", + " 119,\n", + " 119,\n", + " 119,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", + " 118,\n", " 118,\n", " 118,\n", " 118,\n", @@ -3442,10 +3776,22 @@ " 116,\n", " 116,\n", " 116,\n", + " 116,\n", + " 116,\n", + " 116,\n", + " 116,\n", + " 116,\n", + " 115,\n", + " 115,\n", + " 115,\n", " 115,\n", " 115,\n", " 115,\n", " 115,\n", + " 115,\n", + " 115,\n", + " 115,\n", + " 114,\n", " 114,\n", " 114,\n", " 114,\n", @@ -3456,21 +3802,93 @@ " 114,\n", " 114,\n", " 114,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 113,\n", + " 112,\n", + " 112,\n", + " 112,\n", + " 112,\n", + " 112,\n", + " 112,\n", " 112,\n", " 112,\n", " 112,\n", " 111,\n", - " 110,\n", - " 109,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 111,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 110,\n", + " 109,\n", + " 109,\n", + " 109,\n", + " 109,\n", + " 109,\n", + " 109,\n", + " 109,\n", + " 109,\n", + " 109,\n", " 109,\n", " 109,\n", " 108,\n", " 108,\n", " 108,\n", " 108,\n", + " 108,\n", + " 108,\n", + " 108,\n", + " 108,\n", + " 108,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", + " 107,\n", " 107,\n", " 107,\n", " 107,\n", + " 107,\n", + " 107,\n", + " 106,\n", + " 106,\n", + " 106,\n", + " 106,\n", " 106,\n", " 106,\n", " 106,\n", @@ -3483,14 +3901,32 @@ " 105,\n", " 105,\n", " 105,\n", + " 105,\n", + " 105,\n", + " 105,\n", + " 105,\n", + " 105,\n", " 104,\n", " 104,\n", " 104,\n", " 104,\n", " 104,\n", - " 104,\n", " 103,\n", " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 103,\n", + " 102,\n", + " 102,\n", + " 102,\n", + " 102,\n", " 102,\n", " 102,\n", " 102,\n", @@ -3499,6 +3935,15 @@ " 102,\n", " 102,\n", " 102,\n", + " 102,\n", + " 102,\n", + " 102,\n", + " 102,\n", + " 101,\n", + " 101,\n", + " 101,\n", + " 101,\n", + " 101,\n", " 101,\n", " 101,\n", " 101,\n", @@ -3513,15 +3958,61 @@ " 99,\n", " 99,\n", " 99,\n", + " 99,\n", + " 99,\n", + " 99,\n", + " 99,\n", + " 99,\n", + " 99,\n", + " 99,\n", + " 99,\n", + " 98,\n", + " 98,\n", + " 98,\n", + " 98,\n", + " 98,\n", + " 98,\n", + " 98,\n", + " 98,\n", " 98,\n", " 98,\n", " 98,\n", " 98,\n", " 98,\n", " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", + " 97,\n", " 96,\n", " 96,\n", " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 96,\n", + " 95,\n", + " 95,\n", + " 95,\n", " 95,\n", " 95,\n", " 95,\n", @@ -3529,14 +4020,69 @@ " 95,\n", " 95,\n", " 95,\n", + " 95,\n", + " 95,\n", + " 95,\n", + " 94,\n", " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 94,\n", + " 93,\n", + " 93,\n", + " 93,\n", " 93,\n", " 93,\n", " 93,\n", " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 93,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", " 92,\n", " 92,\n", " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 92,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", + " 91,\n", " 91,\n", " 91,\n", " 91,\n", @@ -3550,6 +4096,26 @@ " 90,\n", " 90,\n", " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 90,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", + " 89,\n", " 89,\n", " 89,\n", " 89,\n", @@ -3570,6 +4136,29 @@ " 87,\n", " 87,\n", " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 87,\n", + " 86,\n", + " 86,\n", + " 86,\n", + " 86,\n", + " 86,\n", + " 86,\n", + " 86,\n", " 86,\n", " 86,\n", " 86,\n", @@ -3577,10 +4166,35 @@ " 86,\n", " 86,\n", " 86,\n", + " 86,\n", + " 86,\n", + " 86,\n", + " 86,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", " 85,\n", " 85,\n", " 85,\n", " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 85,\n", + " 84,\n", + " 84,\n", " 84,\n", " 84,\n", " 84,\n", @@ -3588,6 +4202,23 @@ " 84,\n", " 84,\n", " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 84,\n", + " 83,\n", + " 83,\n", + " 83,\n", + " 83,\n", + " 83,\n", + " 83,\n", + " 83,\n", + " 83,\n", " 83,\n", " 83,\n", " 83,\n", @@ -3595,638 +4226,10 @@ " 83,\n", " 83,\n", " 83,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 82,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 81,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 80,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 79,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 78,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 77,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 76,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 75,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 74,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 73,\n", - " 72,\n", - " 72,\n", - " 72,\n", - " 72,\n", - " 72,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 71,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 70,\n", - " 69,\n", - " 69,\n", - " 69,\n", - " 69,\n", - " 69,\n", - " 69,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 68,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 67,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 66,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 65,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 64,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 63,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 62,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 61,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 60,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 59,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 58,\n", - " 57,\n", - " 57,\n", - " 57,\n", - " 57,\n", - " 57,\n", - " 57,\n", - " 57,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 56,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 55,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 54,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 53,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 52,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 51,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 50,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 49,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 48,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 47,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 46,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 45,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 44,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 43,\n", - " 42,\n", - " 42,\n", - " 42,\n", - " 42,\n", " ...]" ] }, - "execution_count": 5, + "execution_count": 10, "metadata": {}, "output_type": "execute_result" } diff --git a/configs/rscup/ga.py b/configs/rscup/ga.py new file mode 100644 index 0000000..8acdbc4 --- /dev/null +++ b/configs/rscup/ga.py @@ -0,0 +1,174 @@ +# model settings +fp16 = dict(loss_scale=512.) +model = dict( + type='RPN', + pretrained='modelzoo://resnet50', + backbone=dict( + type='ResNet', + depth=50, + num_stages=4, + out_indices=(0, 1, 2, 3), + frozen_stages=1, + style='pytorch', + dcn=dict( + modulated=False, deformable_groups=1, fallback_on_stride=False), + stage_with_dcn=(False, True, True, True), + ), + neck=dict( + type='FPN', + in_channels=[256, 512, 1024, 2048], + out_channels=256, + num_outs=5), + rpn_head=dict( + type='GARPNHead', + in_channels=256, + feat_channels=256, + octave_base_scale=8, + scales_per_octave=3, + octave_ratios=[0.5, 1.0, 2.0], + anchor_strides=[4, 8, 16, 32, 64], + anchor_base_sizes=None, + anchoring_means=[.0, .0, .0, .0], + anchoring_stds=[0.07, 0.07, 0.14, 0.14], + target_means=(.0, .0, .0, .0), + target_stds=[0.07, 0.07, 0.11, 0.11], + loc_filter_thr=0.01, + loss_loc=dict( + type='FocalLoss', + use_sigmoid=True, + gamma=2.0, + alpha=0.25, + loss_weight=1.0), + loss_shape=dict(type='BoundedIoULoss', beta=0.2, loss_weight=1.0), + loss_cls=dict( + type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), + loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)) +) +# model training and testing settings +train_cfg = dict( + rpn=dict( + ga_assigner=dict( + type='ApproxMaxIoUAssigner', + pos_iou_thr=0.7, + neg_iou_thr=0.3, + min_pos_iou=0.3, + ignore_iof_thr=-1), + ga_sampler=dict( + type='RandomSampler', + num=512, + pos_fraction=0.5, + neg_pos_ub=-1, + add_gt_as_proposals=False), + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.7, + neg_iou_thr=0.3, + min_pos_iou=0.3, + ignore_iof_thr=-1), + sampler=dict( + type='RandomSampler', + num=512, + pos_fraction=0.5, + neg_pos_ub=-1, + add_gt_as_proposals=False), + allowed_border=-1, + pos_weight=-1, + center_ratio=0.2, + ignore_ratio=0.5, + debug=False), +) +test_cfg = dict( + rpn=dict( + nms_across_levels=False, + nms_pre=1000, + nms_post=1000, + max_num=500, + nms_thr=0.7, + min_bbox_size=0), +) +# dataset settings +dataset_type = 'CocoDataset' +data_root = './data/rscup/' +aug_root = "./data/rscup/aug/" +other_aug_root = "./data/rscup/otheraug/" +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +data = dict( + imgs_per_gpu=6, + workers_per_gpu=4, + train=dict( + type=dataset_type, + ann_file=(data_root + 'annotation/annos_rscup_train.json', + aug_root + 'annos_rscup_airport.json', + other_aug_root + "annos_rscup_baseball-diamond.json", + other_aug_root + "annos_rscup_basketball-court.json", + other_aug_root + "annos_rscup_container-crane.json", + other_aug_root + "annos_rscup_helicopter.json", + other_aug_root + "annos_rscup_helipad.json", + other_aug_root + "annos_rscup_helipad_ship.json", + other_aug_root + "annos_rscup_roundabout.json", + other_aug_root + "annos_rscup_soccer-ball-field_ground-track-field.json", + ), + img_prefix=(data_root + 'train/', + aug_root + "airport/", + other_aug_root + "baseball-diamond", + other_aug_root + "basketball-court", + other_aug_root + "container-crane", + other_aug_root + "helicopter", + other_aug_root + "helipad", + other_aug_root + "helipad_ship", + other_aug_root + "roundabout", + other_aug_root + "soccer-ball-field_ground-track-field"), + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0.5, + with_mask=False, + with_crowd=False, + with_label=False), + val=dict( + type=dataset_type, + ann_file=data_root + 'annotation/annos_rscup_val.json', + img_prefix=data_root + 'val/', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_crowd=True, + with_label=True), + test=dict( + type=dataset_type, + ann_file='./data/rscup/debug.json', + img_prefix='./data/rscup/debug/', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_label=False, + test_mode=True)) +# optimizer +optimizer = dict(type='SGD', lr=3e-3, momentum=0.9, weight_decay=0.0001) +optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2)) +# learning policy +lr_config = dict( + policy='step', + step=[2, 4, 6]) +checkpoint_config = dict(interval=1) +# yapf:disable +log_config = dict( + interval=20, + hooks=[ + dict(type='TextLoggerHook'), + dict(type='TensorboardLoggerHook') + ]) +# yapf:enable +# runtime settings +total_epochs = 8 +dist_params = dict(backend='nccl') +log_level = 'INFO' +work_dir = './work_dirs/ga' +load_from = None +resume_from = None +workflow = [('train', 1)] diff --git a/configs/rscup/htc_anchor.py b/configs/rscup/htc_anchor.py new file mode 100644 index 0000000..bc613ac --- /dev/null +++ b/configs/rscup/htc_anchor.py @@ -0,0 +1,307 @@ +# model settings +fp16 = dict(loss_scale=512.) +# norm_cfg = dict(type='GN', num_groups=32, requires_grad=True) +norm_cfg = dict(type='SyncBN', requires_grad=True) +model = dict( + type='HybridTaskCascade', + num_stages=3, + pretrained='modelzoo://resnet50', + interleaved=True, + mask_info_flow=True, + backbone=dict( + type='ResNet', + depth=50, + num_stages=4, + out_indices=(0, 1, 2, 3), + frozen_stages=1, + style='pytorch', + dcn=dict( + modulated=False, deformable_groups=1, fallback_on_stride=False), + stage_with_dcn=(False, True, True, True), + ), + neck=dict( + type='FPN', + in_channels=[256, 512, 1024, 2048], + out_channels=256, + num_outs=5), + rpn_head=dict( + type='RPNHead', + in_channels=256, + feat_channels=256, + anchor_scales=[8], + anchor_ratios=[0.5, 1.0, 1.5, 2.5, 5.0, 7.0], + anchor_strides=[4, 8, 16, 32, 64], + target_means=[.0, .0, .0, .0], + target_stds=[1.0, 1.0, 1.0, 1.0], + loss_cls=dict( + type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), + loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)), + bbox_roi_extractor=dict( + type='SingleRoIExtractor', + roi_layer=dict(type='RoIAlign', out_size=7, sample_num=2), + out_channels=256, + featmap_strides=[4, 8, 16, 32]), + bbox_head=[ + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.1, 0.1, 0.2, 0.2], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='SmoothL1Loss', + beta=1.0, + loss_weight=1.0)), + + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.05, 0.05, 0.1, 0.1], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='SmoothL1Loss', + beta=1.0, + loss_weight=1.0)), + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.033, 0.033, 0.067, 0.067], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='SmoothL1Loss', + beta=1.0, + loss_weight=1.0)) + ], + mask_roi_extractor=dict( + type='SingleRoIExtractor', + roi_layer=dict(type='RoIAlign', out_size=14, sample_num=2), + out_channels=256, + featmap_strides=[4, 8, 16, 32]), + mask_head=dict( + type='HTCMaskHead', + num_convs=4, + in_channels=256, + conv_out_channels=256, + num_classes=19, + loss_mask=dict( + type='CrossEntropyLoss', use_mask=True, loss_weight=1.0))) +# model training and testing settings +train_cfg = dict( + rpn=dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.7, + neg_iou_thr=0.3, + min_pos_iou=0.3, + ignore_iof_thr=-1), + sampler=dict( + type='RandomSampler', + num=256, + pos_fraction=0.5, + neg_pos_ub=-1, + add_gt_as_proposals=False), + allowed_border=0, + pos_weight=-1, + debug=False), + rpn_proposal=dict( + nms_across_levels=False, + nms_pre=2000, + nms_post=2000, + max_num=2000, + nms_thr=0.7, + min_bbox_size=0), + rcnn=[ + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.4, + neg_iou_thr=0.4, + min_pos_iou=0.4, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False), + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.5, + neg_iou_thr=0.5, + min_pos_iou=0.5, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False), + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.6, + neg_iou_thr=0.6, + min_pos_iou=0.6, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False) + ], + stage_loss_weights=[0.5, 1, 0.25]) +test_cfg = dict( + rpn=dict( + nms_across_levels=False, + nms_pre=1000, + nms_post=1000, + max_num=1000, + nms_thr=0.7, + min_bbox_size=0), + rcnn=dict( + score_thr=0.05, + nms=dict(type='nms', iou_thr=0.5), + max_per_img=1000, + mask_thr_binary=0.5), + keep_all_stages=False) +# dataset settings +dataset_type = 'CocoDataset' +data_root = './data/rscup/' +aug_root = "./data/rscup/aug/" +other_aug_root = "./data/rscup/otheraug/" +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +data = dict( + imgs_per_gpu=6, + workers_per_gpu=4, + train=dict( + type=dataset_type, + ann_file=(data_root + 'annotation/annos_rscup_train.json', + aug_root + 'annos_rscup_airport.json', + other_aug_root + "annos_rscup_baseball-diamond.json", + other_aug_root + "annos_rscup_basketball-court.json", + other_aug_root + "annos_rscup_container-crane.json", + other_aug_root + "annos_rscup_helicopter.json", + other_aug_root + "annos_rscup_helipad.json", + other_aug_root + "annos_rscup_helipad_ship.json", + other_aug_root + "annos_rscup_roundabout.json", + other_aug_root + "annos_rscup_soccer-ball-field_ground-track-field.json", + ), + img_prefix=(data_root + 'train/', + aug_root + "airport/", + other_aug_root + "baseball-diamond", + other_aug_root + "basketball-court", + other_aug_root + "container-crane", + other_aug_root + "helicopter", + other_aug_root + "helipad", + other_aug_root + "helipad_ship", + other_aug_root + "roundabout", + other_aug_root + "soccer-ball-field_ground-track-field"), + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0.5, + with_mask=True, + with_crowd=True, + with_label=True), + val=dict( + type=dataset_type, + ann_file=data_root + 'annotation/annos_rscup_val.json', + img_prefix=data_root + 'val/', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_crowd=True, + with_label=True), + test=dict( + type=dataset_type, + ann_file='./data/rscup/annotation/annos_rscup_test.json', + img_prefix='./data/rscup/test', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_label=False, + test_mode=True)) +# optimizer +optimizer = dict(type='SGD', lr=9e-3, momentum=0.9, weight_decay=0.0001) +optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2)) +# learning policy +lr_config = dict( + policy='step', + warmup='linear', + warmup_iters=500, + warmup_ratio=1.0 / 3, + step=[8, 11]) +checkpoint_config = dict(interval=1) +# yapf:disable +log_config = dict( + interval=50, + hooks=[ + dict(type='TextLoggerHook'), + dict(type='TensorboardLoggerHook') + ]) +# yapf:enable +# runtime settings +total_epochs = 12 +dist_params = dict(backend='nccl') +log_level = 'INFO' +work_dir = './work_dirs/htc_sy' +load_from = None +resume_from = None +workflow = [('train', 1)] diff --git a/configs/rscup/htc_deform_focal.py b/configs/rscup/htc_deform_focal.py index 0fe82c5..8aa7ff0 100644 --- a/configs/rscup/htc_deform_focal.py +++ b/configs/rscup/htc_deform_focal.py @@ -252,8 +252,8 @@ with_label=True), test=dict( type=dataset_type, - ann_file='./data/rscup/annotation/annos_rscup_test.json', - img_prefix='./data/rscup/test', + ann_file='./data/rscup/annotation/annos_rscup_val.json', + img_prefix='./data/rscup/val', img_scale=(512, 512), img_norm_cfg=img_norm_cfg, size_divisor=32, diff --git a/configs/rscup/htc_libra.py b/configs/rscup/htc_libra.py new file mode 100644 index 0000000..9627de7 --- /dev/null +++ b/configs/rscup/htc_libra.py @@ -0,0 +1,313 @@ +# model settings +fp16 = dict(loss_scale=512.) +# norm_cfg = dict(type='GN', num_groups=32, requires_grad=True) +norm_cfg = dict(type='SyncBN', requires_grad=True) +model = dict( + type='HybridTaskCascade', + num_stages=3, + pretrained='modelzoo://resnet50', + interleaved=True, + mask_info_flow=True, + backbone=dict( + type='ResNet', + depth=50, + num_stages=4, + out_indices=(0, 1, 2, 3), + frozen_stages=1, + style='pytorch', + dcn=dict( + modulated=False, deformable_groups=1, fallback_on_stride=False), + stage_with_dcn=(False, True, True, True), + ), + neck=dict( + type='FPN', + in_channels=[256, 512, 1024, 2048], + out_channels=256, + num_outs=5), + rpn_head=dict( + type='RPNHead', + in_channels=256, + feat_channels=256, + anchor_scales=[8], + anchor_ratios=[0.5, 1.0, 1.5, 2.5, 5.0, 7.0], + anchor_strides=[4, 8, 16, 32, 64], + target_means=[.0, .0, .0, .0], + target_stds=[1.0, 1.0, 1.0, 1.0], + loss_cls=dict( + type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), + loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)), + bbox_roi_extractor=dict( + type='SingleRoIExtractor', + roi_layer=dict(type='RoIAlign', out_size=7, sample_num=2), + out_channels=256, + featmap_strides=[4, 8, 16, 32]), + bbox_head=[ + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.1, 0.1, 0.2, 0.2], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='BalancedL1Loss', + alpha=0.5, + gamma=1.5, + beta=1.0, + loss_weight=1.0)), + + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.05, 0.05, 0.1, 0.1], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='BalancedL1Loss', + alpha=0.5, + gamma=1.5, + beta=1.0, + loss_weight=1.0)), + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.033, 0.033, 0.067, 0.067], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='BalancedL1Loss', + alpha=0.5, + gamma=1.5, + beta=1.0, + loss_weight=1.0)) + ], + mask_roi_extractor=dict( + type='SingleRoIExtractor', + roi_layer=dict(type='RoIAlign', out_size=14, sample_num=2), + out_channels=256, + featmap_strides=[4, 8, 16, 32]), + mask_head=dict( + type='HTCMaskHead', + num_convs=4, + in_channels=256, + conv_out_channels=256, + num_classes=19, + loss_mask=dict( + type='CrossEntropyLoss', use_mask=True, loss_weight=1.0))) +# model training and testing settings +train_cfg = dict( + rpn=dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.7, + neg_iou_thr=0.3, + min_pos_iou=0.3, + ignore_iof_thr=-1), + sampler=dict( + type='RandomSampler', + num=256, + pos_fraction=0.5, + neg_pos_ub=-1, + add_gt_as_proposals=False), + allowed_border=0, + pos_weight=-1, + debug=False), + rpn_proposal=dict( + nms_across_levels=False, + nms_pre=2000, + nms_post=2000, + max_num=2000, + nms_thr=0.7, + min_bbox_size=0), + rcnn=[ + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.4, + neg_iou_thr=0.4, + min_pos_iou=0.4, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False), + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.5, + neg_iou_thr=0.5, + min_pos_iou=0.5, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False), + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.6, + neg_iou_thr=0.6, + min_pos_iou=0.6, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False) + ], + stage_loss_weights=[0.5, 1, 0.25]) +test_cfg = dict( + rpn=dict( + nms_across_levels=False, + nms_pre=1000, + nms_post=1000, + max_num=1000, + nms_thr=0.7, + min_bbox_size=0), + rcnn=dict( + score_thr=0.05, + nms=dict(type='nms', iou_thr=0.5), + max_per_img=1000, + mask_thr_binary=0.5), + keep_all_stages=False) +# dataset settings +dataset_type = 'CocoDataset' +data_root = './data/rscup/' +aug_root = "./data/rscup/aug/" +other_aug_root = "./data/rscup/otheraug/" +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +data = dict( + imgs_per_gpu=6, + workers_per_gpu=4, + train=dict( + type=dataset_type, + ann_file=(data_root + 'annotation/annos_rscup_train.json', + aug_root + 'annos_rscup_airport.json', + other_aug_root + "annos_rscup_baseball-diamond.json", + other_aug_root + "annos_rscup_basketball-court.json", + other_aug_root + "annos_rscup_container-crane.json", + other_aug_root + "annos_rscup_helicopter.json", + other_aug_root + "annos_rscup_helipad.json", + other_aug_root + "annos_rscup_helipad_ship.json", + other_aug_root + "annos_rscup_roundabout.json", + other_aug_root + "annos_rscup_soccer-ball-field_ground-track-field.json", + ), + img_prefix=(data_root + 'train/', + aug_root + "airport/", + other_aug_root + "baseball-diamond", + other_aug_root + "basketball-court", + other_aug_root + "container-crane", + other_aug_root + "helicopter", + other_aug_root + "helipad", + other_aug_root + "helipad_ship", + other_aug_root + "roundabout", + other_aug_root + "soccer-ball-field_ground-track-field"), + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0.5, + with_mask=True, + with_crowd=True, + with_label=True), + val=dict( + type=dataset_type, + ann_file=data_root + 'annotation/annos_rscup_val.json', + img_prefix=data_root + 'val/', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_crowd=True, + with_label=True), + test=dict( + type=dataset_type, + ann_file='./data/rscup/annotation/annos_rscup_test.json', + img_prefix='./data/rscup/test', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_label=False, + test_mode=True)) +# optimizer +optimizer = dict(type='SGD', lr=9e-3, momentum=0.9, weight_decay=0.0001) +optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2)) +# learning policy +lr_config = dict( + policy='step', + warmup='linear', + warmup_iters=500, + warmup_ratio=1.0 / 3, + step=[8, 11]) +checkpoint_config = dict(interval=1) +# yapf:disable +log_config = dict( + interval=50, + hooks=[ + dict(type='TextLoggerHook'), + dict(type='TensorboardLoggerHook') + ]) +# yapf:enable +# runtime settings +total_epochs = 12 +dist_params = dict(backend='nccl') +log_level = 'INFO' +work_dir = './work_dirs/htc_libra' +load_from = None +resume_from = None +workflow = [('train', 1)] diff --git a/configs/rscup/htc_sy.py b/configs/rscup/htc_sy.py new file mode 100644 index 0000000..bc613ac --- /dev/null +++ b/configs/rscup/htc_sy.py @@ -0,0 +1,307 @@ +# model settings +fp16 = dict(loss_scale=512.) +# norm_cfg = dict(type='GN', num_groups=32, requires_grad=True) +norm_cfg = dict(type='SyncBN', requires_grad=True) +model = dict( + type='HybridTaskCascade', + num_stages=3, + pretrained='modelzoo://resnet50', + interleaved=True, + mask_info_flow=True, + backbone=dict( + type='ResNet', + depth=50, + num_stages=4, + out_indices=(0, 1, 2, 3), + frozen_stages=1, + style='pytorch', + dcn=dict( + modulated=False, deformable_groups=1, fallback_on_stride=False), + stage_with_dcn=(False, True, True, True), + ), + neck=dict( + type='FPN', + in_channels=[256, 512, 1024, 2048], + out_channels=256, + num_outs=5), + rpn_head=dict( + type='RPNHead', + in_channels=256, + feat_channels=256, + anchor_scales=[8], + anchor_ratios=[0.5, 1.0, 1.5, 2.5, 5.0, 7.0], + anchor_strides=[4, 8, 16, 32, 64], + target_means=[.0, .0, .0, .0], + target_stds=[1.0, 1.0, 1.0, 1.0], + loss_cls=dict( + type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), + loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)), + bbox_roi_extractor=dict( + type='SingleRoIExtractor', + roi_layer=dict(type='RoIAlign', out_size=7, sample_num=2), + out_channels=256, + featmap_strides=[4, 8, 16, 32]), + bbox_head=[ + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.1, 0.1, 0.2, 0.2], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='SmoothL1Loss', + beta=1.0, + loss_weight=1.0)), + + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.05, 0.05, 0.1, 0.1], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='SmoothL1Loss', + beta=1.0, + loss_weight=1.0)), + dict( + type='SharedFCBBoxHead', + num_fcs=2, + in_channels=256, + fc_out_channels=1024, + roi_feat_size=7, + num_classes=19, + target_means=[0., 0., 0., 0.], + target_stds=[0.033, 0.033, 0.067, 0.067], + reg_class_agnostic=True, + loss_cls=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + loss_bbox=dict( + type='SmoothL1Loss', + beta=1.0, + loss_weight=1.0)) + ], + mask_roi_extractor=dict( + type='SingleRoIExtractor', + roi_layer=dict(type='RoIAlign', out_size=14, sample_num=2), + out_channels=256, + featmap_strides=[4, 8, 16, 32]), + mask_head=dict( + type='HTCMaskHead', + num_convs=4, + in_channels=256, + conv_out_channels=256, + num_classes=19, + loss_mask=dict( + type='CrossEntropyLoss', use_mask=True, loss_weight=1.0))) +# model training and testing settings +train_cfg = dict( + rpn=dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.7, + neg_iou_thr=0.3, + min_pos_iou=0.3, + ignore_iof_thr=-1), + sampler=dict( + type='RandomSampler', + num=256, + pos_fraction=0.5, + neg_pos_ub=-1, + add_gt_as_proposals=False), + allowed_border=0, + pos_weight=-1, + debug=False), + rpn_proposal=dict( + nms_across_levels=False, + nms_pre=2000, + nms_post=2000, + max_num=2000, + nms_thr=0.7, + min_bbox_size=0), + rcnn=[ + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.4, + neg_iou_thr=0.4, + min_pos_iou=0.4, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False), + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.5, + neg_iou_thr=0.5, + min_pos_iou=0.5, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False), + dict( + assigner=dict( + type='MaxIoUAssigner', + pos_iou_thr=0.6, + neg_iou_thr=0.6, + min_pos_iou=0.6, + ignore_iof_thr=-1), + sampler=dict( + type='CombinedSampler', + num=512, + pos_fraction=0.25, + add_gt_as_proposals=True, + pos_sampler=dict(type='InstanceBalancedPosSampler'), + neg_sampler=dict( + type='IoUBalancedNegSampler', + floor_thr=-1, + floor_fraction=0, + num_bins=3)), + mask_size=28, + pos_weight=-1, + debug=False) + ], + stage_loss_weights=[0.5, 1, 0.25]) +test_cfg = dict( + rpn=dict( + nms_across_levels=False, + nms_pre=1000, + nms_post=1000, + max_num=1000, + nms_thr=0.7, + min_bbox_size=0), + rcnn=dict( + score_thr=0.05, + nms=dict(type='nms', iou_thr=0.5), + max_per_img=1000, + mask_thr_binary=0.5), + keep_all_stages=False) +# dataset settings +dataset_type = 'CocoDataset' +data_root = './data/rscup/' +aug_root = "./data/rscup/aug/" +other_aug_root = "./data/rscup/otheraug/" +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +data = dict( + imgs_per_gpu=6, + workers_per_gpu=4, + train=dict( + type=dataset_type, + ann_file=(data_root + 'annotation/annos_rscup_train.json', + aug_root + 'annos_rscup_airport.json', + other_aug_root + "annos_rscup_baseball-diamond.json", + other_aug_root + "annos_rscup_basketball-court.json", + other_aug_root + "annos_rscup_container-crane.json", + other_aug_root + "annos_rscup_helicopter.json", + other_aug_root + "annos_rscup_helipad.json", + other_aug_root + "annos_rscup_helipad_ship.json", + other_aug_root + "annos_rscup_roundabout.json", + other_aug_root + "annos_rscup_soccer-ball-field_ground-track-field.json", + ), + img_prefix=(data_root + 'train/', + aug_root + "airport/", + other_aug_root + "baseball-diamond", + other_aug_root + "basketball-court", + other_aug_root + "container-crane", + other_aug_root + "helicopter", + other_aug_root + "helipad", + other_aug_root + "helipad_ship", + other_aug_root + "roundabout", + other_aug_root + "soccer-ball-field_ground-track-field"), + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0.5, + with_mask=True, + with_crowd=True, + with_label=True), + val=dict( + type=dataset_type, + ann_file=data_root + 'annotation/annos_rscup_val.json', + img_prefix=data_root + 'val/', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_crowd=True, + with_label=True), + test=dict( + type=dataset_type, + ann_file='./data/rscup/annotation/annos_rscup_test.json', + img_prefix='./data/rscup/test', + img_scale=(512, 512), + img_norm_cfg=img_norm_cfg, + size_divisor=32, + flip_ratio=0, + with_mask=True, + with_label=False, + test_mode=True)) +# optimizer +optimizer = dict(type='SGD', lr=9e-3, momentum=0.9, weight_decay=0.0001) +optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2)) +# learning policy +lr_config = dict( + policy='step', + warmup='linear', + warmup_iters=500, + warmup_ratio=1.0 / 3, + step=[8, 11]) +checkpoint_config = dict(interval=1) +# yapf:disable +log_config = dict( + interval=50, + hooks=[ + dict(type='TextLoggerHook'), + dict(type='TensorboardLoggerHook') + ]) +# yapf:enable +# runtime settings +total_epochs = 12 +dist_params = dict(backend='nccl') +log_level = 'INFO' +work_dir = './work_dirs/htc_sy' +load_from = None +resume_from = None +workflow = [('train', 1)] diff --git a/generate_chips.ipynb b/generate_chips.ipynb index ba4319b..8a10048 100644 --- a/generate_chips.ipynb +++ b/generate_chips.ipynb @@ -2,8 +2,10 @@ "cells": [ { "cell_type": "code", - "execution_count": 69, - "metadata": {}, + "execution_count": 1, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "import cv2\n", @@ -36,7 +38,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "datadir = \"./data/train/images\"\n", @@ -97,7 +101,9 @@ { "cell_type": "code", "execution_count": 70, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def get_info(fp):\n", @@ -162,7 +168,9 @@ { "cell_type": "code", "execution_count": 39, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def bbox_overlaps_py(boxes, query_boxes):\n", @@ -1714,7 +1722,9 @@ { "cell_type": "code", "execution_count": 307, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def _pygenerate_noobj(boxes, masks, labels, width, height, chipsize, stride):\n", @@ -1964,7 +1974,9 @@ { "cell_type": "code", "execution_count": 44, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def aug_roi(img):\n", @@ -2538,7 +2550,9 @@ { "cell_type": "code", "execution_count": 174, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def generate_objects(classname):\n", @@ -2622,7 +2636,9 @@ { "cell_type": "code", "execution_count": 55, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def merge_object_background(materials, objects, info, source, des):\n", @@ -3542,7 +3558,9 @@ { "cell_type": "code", "execution_count": 235, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [ "def noise_generate(datadir, phase):\n", @@ -4613,7 +4631,9 @@ { "cell_type": "code", "execution_count": null, - "metadata": {}, + "metadata": { + "collapsed": true + }, "outputs": [], "source": [] } diff --git a/mmdet/apis/inference.py b/mmdet/apis/inference.py index cbbd127..e536fbd 100644 --- a/mmdet/apis/inference.py +++ b/mmdet/apis/inference.py @@ -120,6 +120,7 @@ def show_result(img, result, class_names, score_thr=0.3, out_file=None): bbox_result, segm_result = result else: bbox_result, segm_result = result, None + ic(bbox_result) bboxes = np.vstack(bbox_result) # draw segmentation masks if segm_result is not None: @@ -130,11 +131,14 @@ def show_result(img, result, class_names, score_thr=0.3, out_file=None): mask = maskUtils.decode(segms[i]).astype(np.bool) img[mask] = img[mask] * 0.5 + color_mask * 0.5 # draw bounding boxes - labels = [ - np.full(bbox.shape[0], i, dtype=np.int32) - for i, bbox in enumerate(bbox_result) - ] - labels = np.concatenate(labels) + # labels = [ + # np.full(bbox.shape[0], i, dtype=np.int32) + # for i, bbox in enumerate(bbox_result) + # ] + labels = np.array([0 for i in range(len(bbox_result))]) + ic(bboxes.shape) + # labels = np.concatenate(labels) + ic(len(labels)) mmcv.imshow_det_bboxes( img.copy(), bboxes, diff --git a/mmdet/models/anchor_heads/ga_rpn_head.py b/mmdet/models/anchor_heads/ga_rpn_head.py index b7788b6..fed3f50 100644 --- a/mmdet/models/anchor_heads/ga_rpn_head.py +++ b/mmdet/models/anchor_heads/ga_rpn_head.py @@ -67,6 +67,7 @@ def get_bboxes_single(self, cfg, rescale=False): mlvl_proposals = [] + print(cfg.min_bbox_size) for idx in range(len(cls_scores)): rpn_cls_score = cls_scores[idx] rpn_bbox_pred = bbox_preds[idx] diff --git a/mmdet/models/losses/cross_entropy_loss.py b/mmdet/models/losses/cross_entropy_loss.py index 85c4c28..a3b6cfa 100644 --- a/mmdet/models/losses/cross_entropy_loss.py +++ b/mmdet/models/losses/cross_entropy_loss.py @@ -8,11 +8,11 @@ def cross_entropy(pred, label, weight=None, reduction='mean', avg_factor=None): # element-wise losses - W = pred.new(19).fill_(1) - W[11] = 0.75 - W[17] = 0.25 - W[10] = 0.5 - loss = F.cross_entropy(pred, label, weight=W, reduction='none') + # W = pred.new(19).fill_(1) + # W[11] = 0.75 + # W[17] = 0.25 + # W[10] = 0.5 + loss = F.cross_entropy(pred, label, reduction='none') # apply weights and do the reduction if weight is not None: diff --git a/pipeline.ipynb b/pipeline.ipynb index d764378..2101827 100644 --- a/pipeline.ipynb +++ b/pipeline.ipynb @@ -2,10 +2,8 @@ "cells": [ { "cell_type": "code", - "execution_count": 11, - "metadata": { - "collapsed": true - }, + "execution_count": 1, + "metadata": {}, "outputs": [], "source": [ "import pickle\n", @@ -30,10 +28,8 @@ }, { "cell_type": "code", - "execution_count": 12, - "metadata": { - "collapsed": true - }, + "execution_count": 2, + "metadata": {}, "outputs": [], "source": [ "def get_ann(bboxes, segs, cls, name, locx, locy, scale_factor):\n", @@ -133,10 +129,8 @@ }, { "cell_type": "code", - "execution_count": 13, - "metadata": { - "collapsed": true - }, + "execution_count": 3, + "metadata": {}, "outputs": [], "source": [ "def merge_result(config_file, result_file, anno_file, img_prefix, out_file=None, CLASS_NUM=18):\n", @@ -461,10 +455,8 @@ }, { "cell_type": "code", - "execution_count": 14, - "metadata": { - "collapsed": true - }, + "execution_count": 4, + "metadata": {}, "outputs": [], "source": [ "CLASSES = ['tennis-court', 'container-crane', 'storage-tank', 'baseball-diamond', 'plane', 'ground-track-field', 'helicopter', 'airport', 'harbor', 'ship', 'large-vehicle', 'swimming-pool', 'soccer-ball-field', 'roundabout', 'basketball-court', 'bridge', 'small-vehicle', 'helipad']" @@ -479,7 +471,7 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 25, "metadata": {}, "outputs": [ { @@ -487,11 +479,11 @@ "output_type": "stream", "text": [ "loading annotations into memory...\n", - "Done (t=0.25s)\n", + "Done (t=0.27s)\n", "creating index...\n", "index created!\n", "loading annotations into memory...\n", - "Done (t=0.16s)\n", + "Done (t=0.08s)\n", "creating index...\n", "index created!\n" ] @@ -499,7 +491,7 @@ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "a79c506de2de491a9fbc4d0d1397db4a", + "model_id": "837a58fab4934c51985c5fb9588e7088", "version_major": 2, "version_minor": 0 }, @@ -513,7 +505,7 @@ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "138448384cb94781802e61117aba939f", + "model_id": "d1e665a47fe741ca81e67031e71c32bc", "version_major": 2, "version_minor": 0 }, @@ -527,7 +519,7 @@ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "ddf6012d883c4abcb0196823721da110", + "model_id": "38850b3095254be69364748500bb4567", "version_major": 2, "version_minor": 0 }, @@ -543,43 +535,43 @@ "output_type": "stream", "text": [ "tennis-court\n", - "1715\n", + "1641\n", "container-crane\n", - "230\n", + "166\n", "storage-tank\n", - "5307\n", + "5272\n", "baseball-diamond\n", - "1138\n", + "1007\n", "plane\n", - "5145\n", + "5723\n", "ground-track-field\n", - "1301\n", + "1413\n", "helicopter\n", - "368\n", + "369\n", "airport\n", - "172\n", + "166\n", "harbor\n", - "6907\n", + "6916\n", "ship\n", - "24128\n", + "22358\n", "large-vehicle\n", - "20872\n", + "22937\n", "swimming-pool\n", - "2435\n", + "2298\n", "soccer-ball-field\n", - "585\n", + "635\n", "roundabout\n", - "2005\n", + "1780\n", "basketball-court\n", - "810\n", + "770\n", "bridge\n", - "4994\n", + "5071\n", "small-vehicle\n", - "70738\n", + "53677\n", "helipad\n", - "93\n", + "62\n", "loading annotations into memory...\n", - "Done (t=0.67s)\n", + "Done (t=0.76s)\n", "creating index...\n", "index created!\n" ] @@ -587,7 +579,7 @@ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "491e9f68f68c45ce955b2834430a96a0", + "model_id": "ac8a52acd415407a86a8433250c403f6", "version_major": 2, "version_minor": 0 }, @@ -602,31 +594,31 @@ "name": "stdout", "output_type": "stream", "text": [ - "ap of tennis-court is 0.9420828609083332\n", - "ap of container-crane is 0.0\n", - "ap of baseball-diamond is 0.7161250016956516\n", - "ap of ground-track-field is 0.7828051817618804\n", - "ap of storage-tank is 0.6475583369114917\n", - "ap of helicopter is 0.5710499008658795\n", - "ap of airport is 0.7868113379082367\n", - "ap of plane is 0.9034451254980395\n", - "ap of soccer-ball-field is 0.6352499794081782\n", - "ap of swimming-pool is 0.5690378712223216\n", - "ap of roundabout is 0.6445210985182199\n", - "ap of harbor is 0.7411073993923495\n", - "ap of basketball-court is 0.6713363698409788\n", - "ap of helipad is 0.16666666666666666\n", - "ap of bridge is 0.45025108136865716\n", - "ap of large-vehicle is 0.758869159487075\n", - "ap of ship is 0.855791300344259\n", - "ap of small-vehicle is 0.4752380789266801\n", - "map is 0.6287748194847165\n" + "ap of tennis-court is 0.9424778881995032\n", + "ap of container-crane is 0.03826585397333197\n", + "ap of baseball-diamond is 0.7177681700926734\n", + "ap of ground-track-field is 0.7938305490042241\n", + "ap of storage-tank is 0.6472659062184729\n", + "ap of helicopter is 0.4568732141436756\n", + "ap of airport is 0.8265204840847956\n", + "ap of plane is 0.9157028043556301\n", + "ap of swimming-pool is 0.5720345885379551\n", + "ap of soccer-ball-field is 0.6240614306347341\n", + "ap of harbor is 0.7581349170640616\n", + "ap of roundabout is 0.6464088098417231\n", + "ap of basketball-court is 0.6673074170296184\n", + "ap of helipad is 0.125\n", + "ap of bridge is 0.4678420506961264\n", + "ap of large-vehicle is 0.7637899122542678\n", + "ap of ship is 0.8545456046773818\n", + "ap of small-vehicle is 0.4650249712217823\n", + "map is 0.6268252540016642\n" ] } ], "source": [ "config_file = \"configs/rs_cascade_mask_rcnn_r50_fpn_ohem.py\"\n", - "result_file = \"./temp.pkl\"\n", + "result_file = \"./result/val.pkl\"\n", "anno_file = \"/home/xfr/mmdetection/data/rscup/annotation/annos_rscup_val.json\"\n", "out_file = \"./result/eval_temp.pkl\"\n", "img_prefix = \"./data/rscup/val/\"\n", @@ -773,7 +765,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 9, "metadata": {}, "outputs": [ { @@ -785,7 +777,7 @@ "creating index...\n", "index created!\n", "loading annotations into memory...\n", - "Done (t=0.13s)\n", + "Done (t=0.14s)\n", "creating index...\n", "index created!\n" ] @@ -793,7 +785,7 @@ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "e8f16d96833f4c888081ca2ec2de8df5", + "model_id": "ee6e194102504154b351e26cb723652d", "version_major": 2, "version_minor": 0 }, @@ -816,13 +808,13 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "b461ef2a166847fa87f542d2a3f5b446", + "model_id": "c5b617404d5e4f32869143f3d2feac19", "version_major": 2, "version_minor": 0 }, @@ -836,7 +828,7 @@ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "afc9b535bc65424dbe5d15e90d064e88", + "model_id": "77b6b62ea2154628bd25b7e4ba254054", "version_major": 2, "version_minor": 0 }, @@ -852,41 +844,41 @@ "output_type": "stream", "text": [ "tennis-court\n", - "2396\n", + "3035\n", "container-crane\n", - "1996\n", + "2386\n", "storage-tank\n", - "13247\n", + "12862\n", "baseball-diamond\n", - "928\n", + "1169\n", "plane\n", - "2773\n", + "2334\n", "ground-track-field\n", - "1821\n", + "2489\n", "helicopter\n", - "150\n", + "193\n", "airport\n", - "911\n", + "919\n", "harbor\n", - "6487\n", + "8156\n", "ship\n", - "34723\n", + "38982\n", "large-vehicle\n", - "13078\n", + "16052\n", "swimming-pool\n", - "2878\n", + "3614\n", "soccer-ball-field\n", - "1178\n", + "1572\n", "roundabout\n", - "4086\n", + "3944\n", "basketball-court\n", - "1709\n", + "1759\n", "bridge\n", - "9126\n", + "11092\n", "small-vehicle\n", - "53164\n", + "56205\n", "helipad\n", - "824\n" + "558\n" ] } ], @@ -905,13 +897,13 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 11, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { - "model_id": "5e1958ec16d84c75967bb32d69d32513", + "model_id": "ab24e57f8c4c41ffaf4253522952ec54", "version_major": 2, "version_minor": 0 }, @@ -927,41 +919,41 @@ "output_type": "stream", "text": [ "tennis-court\n", - "2396\n", + "3035\n", "container-crane\n", - "1996\n", + "2386\n", "storage-tank\n", - "13247\n", + "12862\n", "baseball-diamond\n", - "928\n", + "1169\n", "plane\n", - "2773\n", + "2334\n", "ground-track-field\n", - "1821\n", + "2489\n", "helicopter\n", - "150\n", + "193\n", "airport\n", - "911\n", + "919\n", "harbor\n", - "6487\n", + "8156\n", "ship\n", - "34723\n", + "38982\n", "large-vehicle\n", - "13078\n", + "16052\n", "swimming-pool\n", - "2878\n", + "3614\n", "soccer-ball-field\n", - "1178\n", + "1572\n", "roundabout\n", - "4086\n", + "3944\n", "basketball-court\n", - "1709\n", + "1759\n", "bridge\n", - "9126\n", + "11092\n", "small-vehicle\n", - "53164\n", + "56205\n", "helipad\n", - "824\n" + "558\n" ] } ], diff --git a/tools/demo.py b/tools/demo.py index 68ecad7..bffdced 100644 --- a/tools/demo.py +++ b/tools/demo.py @@ -22,17 +22,10 @@ def main(): checkpoint_file = args.checkpoint model = init_detector(config_file, checkpoint_file) print(model.CLASSES) - img = './data/rscup/debug/6.jpg' + img = './result/demo/7.jpg' result = inference_detector(model, img) - for name, param in model.named_parameters(): - if name.split('.')[0] != 'rpn_head': - param.requires_grad = False - if param.requires_grad: - print("requires_grad: True ", name) - else: - print("requires_grad: False ", name) - savename = "./result/demo/pic_det2.png" - show_result(img, result, model.CLASSES, score_thr=0.3, out_file=savename) + savename = "./result/demo/pic_det7.png" + show_result(img, result, ['o'], score_thr=0.0, out_file=savename) if __name__ == '__main__': main() \ No newline at end of file