File size: 44,658 Bytes
8aabba6 8d93a57 8aabba6 84afe80 8aabba6 84afe80 8aabba6 8d93a57 8aabba6 8d93a57 8aabba6 8d93a57 8aabba6 8d93a57 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 1a36f9f 8aabba6 8d93a57 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 | {
"meta": {
"title": "SOMA 手部数据图谱",
"updated": "2026-11-02",
"goal": "从单目 RGB 恢复 SOMA hand 的 shape 与 pose：按数据来源和标注方法整理可用的手部数据集。",
"sources": [
{
"id": "ego",
"name": "第一视角",
"desc": "头戴或胸前相机拍的真实视频"
},
{
"id": "egoexo",
"name": "第一视角 + 外视角",
"desc": "同步的头戴相机和固定相机"
},
{
"id": "exo",
"name": "外视角多相机",
"desc": "棚内固定多相机，没有头戴视角"
},
{
"id": "scan",
"name": "扫描仪",
"desc": "静态手势的多相机三维扫描"
},
{
"id": "wild",
"name": "野外图像 / 视频",
"desc": "网络或日常拍摄，没有标定装置"
},
{
"id": "synth",
"name": "合成渲染",
"desc": "渲染或重光照生成的图像"
},
{
"id": "norgb",
"name": "无 RGB",
"desc": "只有动捕、事件相机或黑白相机"
}
],
"methods": [
{
"id": "scan",
"name": "三维扫描配准",
"short": "扫描",
"steps": [
"多相机扫描仪拍静态手势",
"重建亚毫米级网格",
"关键点三角化",
"配准 MANO"
],
"trust": "最高。直接测得手的表面，是目前最可靠的 shape 真值。",
"issues": "只有静态手势，没有手物交互，也没有第一视角。",
"route": "R6"
},
{
"id": "marker",
"name": "光学 marker 动捕 + 拟合",
"short": "光学动捕",
"steps": [
"手上贴反光 marker",
"Vicon / OptiTrack 跟踪",
"逐人标定 shape",
"拟合 MANO 或 SMPL-X"
],
"trust": "高。shape 逐人标定，pose 不受相机遮挡影响。",
"issues": "marker 会出现在 RGB 里；人数通常很少；重遮挡时 marker 丢失。",
"route": "R1"
},
{
"id": "rgbd",
"name": "多视角 RGB-D 拟合",
"short": "RGB-D 拟合",
"steps": [
"多台 RGB-D 同步拍摄",
"2D 关键点（人工或模型）+ 深度",
"多视角联合优化",
"拟合 MANO，人工核验"
],
"trust": "中。受深度噪声和 2D 精度限制，shape 可用但偏粗。",
"issues": "物体遮挡时手指误差大；HOnnotate 一类方法依赖初始化。",
"route": "R1"
},
{
"id": "mvkp",
"name": "多视角关键点三角化",
"short": "关键点三角化",
"steps": [
"多相机 RGB",
"2D 关键点（人工或模型）",
"三角化得到 3D 关键点",
"可选：拟合 MANO"
],
"trust": "中低。shape 只由关键点约束；只给关键点时没有 shape。",
"issues": "模型预标注会带入 HaMeR 等方法的偏差。",
"route": "R1 / R3"
},
{
"id": "device",
"name": "设备追踪",
"short": "设备追踪",
"steps": [
"头显、手套或 IMU",
"设备内置手部追踪",
"输出关节或设备自带手模型"
],
"trust": "低。精度依赖设备，通常没有公开误差。",
"issues": "关节定义因设备而异，需要逐个对齐。",
"route": "R3"
},
{
"id": "synth",
"name": "合成渲染",
"short": "合成",
"steps": [
"参数化手 + 动作库",
"渲染（光照、材质、相机）",
"直接导出真值"
],
"trust": "真值精确，但 shape 分布取决于采样方式。",
"issues": "外观和光照与真实图像有域差。",
"route": "R1 / R2"
},
{
"id": "manual2d",
"name": "人工 2D 标注",
"short": "人工 2D",
"steps": [
"图像",
"模型预标注（可选）",
"人工标注或修正 2D 关键点"
],
"trust": "没有 3D 信息，只提供 2D 监督。",
"issues": "遮挡点的标注口径不一。",
"route": "R4"
},
{
"id": "pseudo",
"name": "单目模型伪标签",
"short": "伪标签",
"steps": [
"单目视频",
"HaMeR 或自研管线预测 MANO",
"可选：修复与筛选"
],
"trust": "最低。不能当 shape 真值，只适合弱监督或预训练。",
"issues": "继承预测模型自身的系统偏差。",
"route": "R7"
}
],
"tiers": [
{
"id": "S1",
"name": "扫描配准",
"desc": "直接测得表面"
},
{
"id": "S2",
"name": "动捕逐人标定",
"desc": "marker 约束的逐人 shape"
},
{
"id": "S3",
"name": "RGB-D 拟合",
"desc": "深度约束，偏粗"
},
{
"id": "S4",
"name": "关键点拟合",
"desc": "只靠关键点约束"
},
{
"id": "S5",
"name": "伪标签",
"desc": "不可作为 shape 真值"
},
{
"id": "SY",
"name": "合成精确",
"desc": "精确，但分布看采样"
},
{
"id": "-",
"name": "无 shape",
"desc": "只有关节或 2D"
}
],
"statuses": [
{
"id": "used",
"name": "已在使用",
"desc": "九源、WiLoR 14 源或基准中已经用到"
},
{
"id": "surveyed",
"name": "团队已调研",
"desc": "已在团队图册中，尚未进训练"
},
{
"id": "retired",
"name": "不再使用",
"desc": "当前不再使用"
},
{
"id": "p0",
"name": "P0 新增",
"desc": "建议优先接入"
},
{
"id": "p1",
"name": "P1 新增",
"desc": "P0 之后接入"
},
{
"id": "p2",
"name": "P2 新增",
"desc": "有余力时接入"
},
{
"id": "own",
"name": "自采",
"desc": "我们自己的采集"
},
{
"id": "no",
"name": "不建议",
"desc": "不满足真实 RGB 或真值条件"
}
],
"routes": [
{
"id": "R0",
"name": "原生 SOMA",
"from": "SOMA 拟合结果",
"how": "直接使用。",
"shape": true
},
{
"id": "R1",
"name": "MANO → SOMA",
"from": "MANO 参数或网格",
"how": "用 SOMAHandLayer 的 MANO 后端映射到 SOMA 拓扑，再重拟合 SOMA 原生 20 维 identity 和 pose；按数据集统计顶点残差，残差大的帧不监督 shape。",
"shape": true
},
{
"id": "R2",
"name": "SMPL-X → SOMA",
"from": "SMPL-X / SMPL-H",
"how": "取手部 MANO 部分走 R1；或用 SOMA-X 的 SMPL-X 身体后端整体转换。手型跟随身体 shape，分布偏窄。",
"shape": true
},
{
"id": "R3",
"name": "3D 关键点 → SOMA pose",
"from": "21 / 25 点 3D 关节",
"how": "先定 MANO 21 点（指尖为网格顶点）与 SOMA 关节的对应，只拟合 pose；shape 固定，不参与监督。",
"shape": false
},
{
"id": "R4",
"name": "2D 关键点",
"from": "2D 关键点",
"how": "不转换。训练时用重投影损失监督。",
"shape": false
},
{
"id": "R5",
"name": "MHR → SOMA",
"from": "MHR 参数",
"how": "SOMAHandLayer 原生支持 MHR 后端，可直接映射。SAM 3D Body 输出 MHR，可当野外图像的伪标签教师。",
"shape": true
},
{
"id": "R6",
"name": "扫描 → SOMA",
"from": "扫描网格",
"how": "直接把 SOMA 原生 shape + pose 拟合到扫描网格，是最好的 shape 监督和评测来源。",
"shape": true
},
{
"id": "R7",
"name": "伪标签",
"from": "预测的 MANO",
"how": "可走 R1 转换，但只作弱监督或预训练，不监督 shape。",
"shape": false
}
],
"teamLinks": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
},
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
},
{
"label": "团队图册：2D 数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_2d_dataset_browser_20261013/index.html"
},
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
]
},
"datasets": [
{
"id": "arctic",
"name": "ARCTIC",
"year": "CVPR 2023",
"status": "used",
"tags": [
"九源",
"WiLoR 14 源",
"基准"
],
"source": "egoexo",
"method": "marker",
"gt": "MANO + SMPL-X",
"tier": "S2",
"route": "R1",
"camera": "8 个固定 RGB + 1 个第一视角 RGB",
"rig": {
"ego": true,
"exo": 8
},
"subjects": "10 人",
"scale": "约 210 万张图，339 个序列",
"fills": "双手操作铰接物体，带第一视角和光学真值。",
"risk": "人数少，shape 分布窄。",
"links": [
{
"label": "论文 arXiv:2204.13662",
"url": "https://arxiv.org/abs/2204.13662"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
},
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
]
},
{
"id": "hot3d",
"name": "HOT3D",
"year": "CVPR 2025",
"status": "used",
"tags": [
"九源",
"WiLoR 14 源",
"基准"
],
"source": "ego",
"method": "marker",
"gt": "MANO / UmeTrack",
"tier": "S2",
"route": "R1",
"camera": "Aria RGB 鱼眼 + Quest 3 黑白",
"rig": {
"ego": true,
"fisheye": true
},
"subjects": "19 人",
"scale": "370 万张以上，约 833 分钟",
"fills": "真实第一视角鱼眼数据，光学动捕真值。",
"risk": "我们的模型在这里 RR 偏差大，先核对裁剪和去畸变与相机模型是否一致。",
"links": [
{
"label": "项目主页",
"url": "https://facebookresearch.github.io/hot3d/"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
},
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
]
},
{
"id": "h2o",
"name": "H2O",
"year": "ICCV 2021",
"status": "used",
"tags": [
"九源",
"基准"
],
"source": "egoexo",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "5 台 Azure Kinect RGB-D，其中 1 台头戴",
"rig": {
"ego": true,
"exo": 4,
"depth": true
},
"subjects": "4 人",
"scale": "571,645 张图，184 个序列",
"fills": "头戴相机与 Femto Bolt 同一传感器族，相机域最接近我们的采集。",
"risk": "人数很少。",
"links": [
{
"label": "论文 arXiv:2104.11181",
"url": "https://arxiv.org/abs/2104.11181"
}
],
"team": [
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
]
},
{
"id": "hoi4d",
"name": "HOI4D",
"year": "CVPR 2022",
"status": "retired",
"tags": [
"九源"
],
"source": "ego",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "头戴 RGB-D",
"rig": {
"ego": true,
"depth": true
},
"subjects": "9 人",
"scale": "约 240 万帧，4,000 个序列",
"fills": "第一视角的类别级手物交互。",
"risk": "右手偏置明显，左手只占约 10%。",
"links": [
{
"label": "论文 arXiv:2203.01577",
"url": "https://arxiv.org/abs/2203.01577"
}
],
"team": [
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
]
},
{
"id": "dexycb",
"name": "DexYCB",
"year": "CVPR 2021",
"status": "used",
"tags": [
"九源",
"WiLoR 14 源"
],
"source": "exo",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "8 台 RGB-D",
"rig": {
"exo": 8,
"depth": true
},
"subjects": "10 人",
"scale": "约 58.2 万张图，1,000 个序列",
"fills": "抓取 YCB 物体，人工多视角 2D 加深度拟合。",
"links": [
{
"label": "论文 arXiv:2104.04631",
"url": "https://arxiv.org/abs/2104.04631"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
},
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
],
"risk": ""
},
{
"id": "ho3d",
"name": "HO3D v3",
"year": "CVPR 2020",
"status": "used",
"tags": [
"九源",
"WiLoR 14 源"
],
"source": "exo",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "1 到 5 台 RGB-D",
"rig": {
"exo": 5,
"depth": true
},
"subjects": "10 人",
"scale": "103,462 张图",
"fills": "单手与物体交互，HOnnotate 联合拟合。",
"links": [
{
"label": "论文 arXiv:1907.01481",
"url": "https://arxiv.org/abs/1907.01481"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
},
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
],
"risk": ""
},
{
"id": "h2o3d",
"name": "H2O3D",
"year": "2022",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "exo",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "5 台 RGB-D",
"rig": {
"exo": 5,
"depth": true
},
"subjects": "待核",
"scale": "76,340 张图，85 个序列",
"fills": "双手与物体交互。",
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"links": [],
"risk": ""
},
{
"id": "interhand",
"name": "InterHand2.6M",
"year": "ECCV 2020",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "exo",
"method": "mvkp",
"gt": "MANO",
"tier": "S4",
"route": "R1",
"camera": "棚内多相机 RGB",
"rig": {
"exo": 12
},
"subjects": "待核",
"scale": "5 fps 版 2,590,347 张",
"fills": "双手交互和丰富的手指姿态。",
"links": [
{
"label": "论文 arXiv:2008.09309",
"url": "https://arxiv.org/abs/2008.09309"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"risk": ""
},
{
"id": "reinterhand",
"name": "Re:InterHand",
"year": "NeurIPS 2023",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "synth",
"method": "mvkp",
"gt": "MANO",
"tier": "S4",
"route": "R1",
"camera": "真实动作重光照渲染，含第一视角",
"rig": {
"synth": true,
"ego": true
},
"scale": "1,526,795 张图，有效 147,745 帧",
"fills": "真实双手动作配多样光照，含第一视角渲染。",
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"links": [],
"risk": "",
"subjects": ""
},
{
"id": "freihand",
"name": "FreiHAND",
"year": "ICCV 2019",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "exo",
"method": "mvkp",
"gt": "MANO",
"tier": "S4",
"route": "R1",
"camera": "棚内多相机，绿幕换背景",
"rig": {
"exo": 8
},
"subjects": "32 人",
"scale": "训练 130,240 张，测试 3,960 张",
"fills": "单手姿态与 shape 的经典基准。",
"links": [
{
"label": "论文 arXiv:1909.04349",
"url": "https://arxiv.org/abs/1909.04349"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"risk": ""
},
{
"id": "mtc",
"name": "MTC",
"year": "CVPR 2019",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "exo",
"method": "mvkp",
"gt": "3D 关键点",
"tier": "-",
"route": "R3",
"camera": "Panoptic 棚内多相机",
"rig": {
"exo": 12
},
"subjects": "40 人",
"scale": "清洗后约 11.1 万个手样本",
"fills": "全身动作中的手，视角多样。",
"links": [
{
"label": "论文 arXiv:1812.01598",
"url": "https://arxiv.org/abs/1812.01598"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"risk": ""
},
{
"id": "bedlam",
"name": "BEDLAM",
"year": "CVPR 2023",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "synth",
"method": "synth",
"gt": "SMPL-X",
"tier": "SY",
"route": "R2",
"camera": "合成渲染，非第一视角",
"rig": {
"synth": true
},
"scale": "10,450 个序列，约 160 万张，约 15 小时",
"fills": "全身场景中的手，背景与光照多样。",
"risk": "手型跟随 SMPL-X 身体 shape，分布偏窄。",
"links": [
{
"label": "论文 arXiv:2306.16940",
"url": "https://arxiv.org/abs/2306.16940"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"subjects": ""
},
{
"id": "rhd",
"name": "RHD",
"year": "ICCV 2017",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "synth",
"method": "synth",
"gt": "3D 关键点",
"tier": "-",
"route": "R3",
"camera": "合成渲染",
"rig": {
"synth": true
},
"scale": "训练 41,258 张，测试 2,728 张",
"fills": "早期合成手部关键点数据。",
"links": [
{
"label": "论文 arXiv:1705.01389",
"url": "https://arxiv.org/abs/1705.01389"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"risk": "",
"subjects": ""
},
{
"id": "egorenderer",
"name": "EgoRenderer V005 / V006",
"year": "2026",
"status": "used",
"tags": [
"自有合成"
],
"source": "synth",
"method": "synth",
"gt": "原生手部参数；拟合 MANO / SOMA 为非 GT 辅助标注",
"tier": "SY",
"route": "R1",
"camera": "第一视角合成渲染",
"rig": {
"synth": true,
"ego": true
},
"scale": "V005：1,314 源组 / 2,425 历史包；V006 ARCTIC：301 源动作，每条只生产 1 份",
"fills": "可控的第一视角合成数据。",
"risk": "V005 参数附件更新沿用原 RGB；V006 渲染使用原生手，不使用拟合参数生成像素。当前只保留每动作一份生产目标，历史额外变体不计入目标。辅助拟合与训练资格另行登记。",
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"links": [],
"subjects": ""
},
{
"id": "cocowb",
"name": "COCO-WholeBody",
"year": "ECCV 2020",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "wild",
"method": "manual2d",
"gt": "2D 关键点",
"tier": "-",
"route": "R4",
"camera": "网络图像",
"rig": {
"wild": true
},
"scale": "约 20 万张图",
"fills": "野外图像的 2D 监督。",
"links": [
{
"label": "论文 arXiv:2007.11858",
"url": "https://arxiv.org/abs/2007.11858"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"risk": "",
"subjects": ""
},
{
"id": "halpe",
"name": "Halpe",
"year": "2020",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "wild",
"method": "manual2d",
"gt": "2D 关键点",
"tier": "-",
"route": "R4",
"camera": "网络图像",
"rig": {
"wild": true
},
"scale": "约 4 万张训练图",
"fills": "野外图像的 2D 监督。",
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"links": [],
"risk": "",
"subjects": ""
},
{
"id": "mpiinzsl",
"name": "MPII + NZSL",
"year": "CVPR 2017",
"status": "used",
"tags": [
"WiLoR 14 源"
],
"source": "wild",
"method": "manual2d",
"gt": "2D 关键点",
"tier": "-",
"route": "R4",
"camera": "网络图像与手语视频",
"rig": {
"wild": true
},
"scale": "2,800 只人工标注的手",
"fills": "人工 2D 手部关键点。",
"links": [
{
"label": "论文 arXiv:1704.07809",
"url": "https://arxiv.org/abs/1704.07809"
}
],
"team": [
{
"label": "团队图册：WiLoR 14 源 + 合成数据",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_dataset_browser_20261013/index.html"
}
],
"risk": "",
"subjects": ""
},
{
"id": "ego4d",
"name": "Ego4D（P3R3 伪标签）",
"year": "CVPR 2022",
"status": "used",
"tags": [
"九源（伪标签）"
],
"source": "wild",
"method": "pseudo",
"gt": "预测 MANO",
"tier": "S5",
"route": "R7",
"camera": "多种头戴相机",
"rig": {
"ego": true,
"wild": true
},
"fills": "海量真实第一视角外观。",
"risk": "P3R3 管线预测的 MANO，不是训练真值。",
"team": [
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
],
"links": [],
"subjects": "",
"scale": ""
},
{
"id": "epic",
"name": "EPIC-KITCHENS（P3R3 伪标签）",
"year": "2018 / 2022",
"status": "used",
"tags": [
"九源（伪标签）"
],
"source": "wild",
"method": "pseudo",
"gt": "预测 MANO",
"tier": "S5",
"route": "R7",
"camera": "头戴 GoPro",
"rig": {
"ego": true,
"wild": true
},
"fills": "厨房场景的真实第一视角。",
"risk": "P3R3 管线预测的 MANO，不是训练真值。",
"team": [
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
],
"links": [],
"subjects": "",
"scale": ""
},
{
"id": "egodex",
"name": "EgoDex",
"year": "2025",
"status": "used",
"tags": [
"九源（伪标签）",
"第一人称图册"
],
"source": "ego",
"method": "device",
"gt": "25 个 3D 关节（ARKit）",
"tier": "-",
"route": "R3",
"camera": "Apple Vision Pro",
"rig": {
"ego": true
},
"scale": "约 829 小时",
"fills": "大规模第一视角灵巧操作，带设备追踪的关节。",
"risk": "关节来自设备追踪，团队另有 P3R3 预测 MANO，都不能监督 shape。",
"links": [
{
"label": "论文 arXiv:2505.11709",
"url": "https://arxiv.org/abs/2505.11709"
}
],
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
},
{
"label": "团队视频：九源 MANO GT 叠加（20s）",
"url": "https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s/tree/main"
}
],
"subjects": ""
},
{
"id": "egopressure",
"name": "EgoPressure",
"year": "CVPR 2025",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "egoexo",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "头戴相机 + 多台外视角 RGB-D",
"rig": {
"ego": true,
"exo": 7,
"depth": true
},
"fills": "手与表面接触的压力，带第一视角。",
"risk": "HaMeR 初始化后用多视角 RGB-D、分割、压力和时序约束拟合 MANO，手上没有 marker。",
"links": [
{
"label": "论文 arXiv:2409.02224",
"url": "https://arxiv.org/abs/2409.02224"
}
],
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"subjects": "",
"scale": ""
},
{
"id": "egoemg",
"name": "EgoEMG",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "marker",
"gt": "MANO（光捕 marker 经 markers2mano 拟合）",
"tier": "S2",
"route": "R1",
"camera": "头戴 GoPro 60 FPS + 26 相机光学动捕",
"rig": {
"ego": true
},
"fills": "第一视角与肌电同步，光学动捕真值。",
"links": [
{
"label": "论文 arXiv:2605.05712",
"url": "https://arxiv.org/abs/2605.05712"
}
],
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"risk": "",
"subjects": "",
"scale": ""
},
{
"id": "hocap",
"name": "HO-Cap",
"year": "2024",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "egoexo",
"method": "rgbd",
"gt": "MANO",
"tier": "S3",
"route": "R1",
"camera": "多台 RGB-D + HoloLens 第一视角 RGB",
"rig": {
"ego": true,
"exo": 8,
"depth": true
},
"fills": "手物交互，含第一视角。",
"risk": "多视角 RGB-D 与关键点约束拟合 MANO，不使用光捕。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"subjects": "",
"scale": ""
},
{
"id": "taco",
"name": "TACO",
"year": "CVPR 2024",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "egoexo",
"method": "mvkp",
"gt": "MANO",
"tier": "S4",
"route": "R1",
"camera": "多台外视角 + 1 个第一视角",
"rig": {
"ego": true,
"exo": 12
},
"fills": "双手使用工具的交互。",
"risk": "多视角 MMPose 三角化后做 MANO 接触与时序拟合，手上没有光捕。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"subjects": "",
"scale": ""
},
{
"id": "aceda0",
"name": "ACE-Data-0",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "egoexo",
"method": "marker",
"gt": "手部动捕（格式待核）",
"tier": "S2",
"route": "R1",
"camera": "环境多相机 + 第一视角",
"rig": {
"ego": true,
"exo": 6
},
"fills": "身体与物体用 OptiTrack，手部附带动捕数据。",
"risk": "官方叠加样例未逐帧配对；全量发布目标 150 小时，目前为快照。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"subjects": "",
"scale": ""
},
{
"id": "xperience",
"name": "Xperience-10M",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "device",
"gt": "手部动捕（类型待核）",
"tier": "-",
"route": "R3",
"camera": "第一视角",
"rig": {
"ego": true
},
"scale": "约 1 万小时",
"fills": "规模极大的第一视角数据。",
"risk": "手部真值的采集方式和精度待核。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"subjects": ""
},
{
"id": "feel",
"name": "FEEL",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "device",
"gt": "Aria MPS 手部追踪",
"tier": "-",
"route": "R3",
"camera": "Aria",
"rig": {
"ego": true,
"fisheye": true
},
"fills": "Aria 第一视角的设备追踪手部。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"risk": "",
"subjects": "",
"scale": ""
},
{
"id": "openaoe",
"name": "Open-AoE",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "pseudo",
"gt": "自动重建 MANO",
"tier": "S5",
"route": "R7",
"camera": "第一视角",
"rig": {
"ego": true
},
"fills": "第一视角操作，自动 MANO 重建。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"risk": "",
"subjects": "",
"scale": ""
},
{
"id": "genhumanego",
"name": "Gen-HumanEgo",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "pseudo",
"gt": "DFM HandTrack 预测",
"tier": "S5",
"route": "R7",
"camera": "第一视角",
"rig": {
"ego": true
},
"fills": "第一视角，模型预测的手部。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"risk": "",
"subjects": "",
"scale": ""
},
{
"id": "egolive",
"name": "EgoLive",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "pseudo",
"gt": "HaMeR 预测 MANO",
"tier": "S5",
"route": "R7",
"camera": "第一视角",
"rig": {
"ego": true
},
"fills": "第一视角，HaMeR 伪标签。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"risk": "",
"subjects": "",
"scale": ""
},
{
"id": "egoverse",
"name": "EgoVerse",
"year": "2026",
"status": "surveyed",
"tags": [
"第一人称图册"
],
"source": "ego",
"method": "device",
"gt": "21 个 3D 关节",
"tier": "-",
"route": "R3",
"camera": "第一视角",
"rig": {
"ego": true
},
"fills": "多地采集的第一视角人类示范。",
"risk": "关节来源待核。",
"links": [
{
"label": "论文 arXiv:2604.07607",
"url": "https://arxiv.org/abs/2604.07607"
}
],
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"subjects": "",
"scale": ""
},
{
"id": "ipnhands",
"name": "IPN HandS",
"year": "2020",
"status": "surveyed",
"tags": [
"2D 图册"
],
"source": "exo",
"method": "manual2d",
"gt": "2D 关键点（21 点）",
"tier": "-",
"route": "R4",
"camera": "单相机",
"rig": {
"exo": 1
},
"scale": "200 段视频，约 80 万帧",
"fills": "MediaPipe 预标后人工修正的连续 2D。",
"risk": "GT 配对和元数据校验未完成。",
"team": [
{
"label": "团队图册：2D 数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/hand_2d_dataset_browser_20261013/index.html"
}
],
"links": [],
"subjects": ""
},
{
"id": "hograspnet",
"name": "HOGraspNet",
"year": "ECCV 2024",
"status": "p0",
"tags": [
"新增"
],
"source": "exo",
"method": "rgbd",
"gt": "MANO + HALO",
"tier": "S3",
"route": "R1",
"camera": "4 台 Azure Kinect RGB-D，其中 1 台在身后眼睛高度模拟第一视角",
"rig": {
"exo": 4,
"depth": true,
"pseudoEgo": true
},
"subjects": "99 人，10 到 74 岁",
"scale": "1,489,112 帧，28 类抓取，30 个物体",
"fills": "shape 多样性：作者的 t-SNE 显示比 HO3D、DexYCB、OakInk 更分散。相机与 Femto Bolt 同一传感器族。",
"risk": "手上没有 marker，MANO 来自 MediaPipe 三角化加 RGB-D 拟合。",
"links": [
{
"label": "项目主页",
"url": "https://hograspnet2024.github.io/"
},
{
"label": "论文 arXiv:2409.04033",
"url": "https://arxiv.org/abs/2409.04033"
}
],
"team": []
},
{
"id": "palm",
"name": "PALM",
"year": "2025",
"status": "p0",
"tags": [
"新增"
],
"source": "scan",
"method": "scan",
"gt": "扫描网格 + MANO 配准",
"tier": "S1",
"route": "R6",
"camera": "3dMD 扫描仪：7 台 RGB + 14 台黑白",
"rig": {
"scanner": true
},
"subjects": "263 人，男女各半，21 到 70 岁",
"scale": "约 9 万张多视角 RGB，1.3 万次扫描，约 50 种右手手势",
"fills": "目前能找到的最可靠 shape 真值，可做 shape 先验和 shape 评测集。",
"risk": "作者称发表后公开；Voxel51 上有 1000 样本子集，完整版获取方式待确认。只有右手静态手势。",
"links": [
{
"label": "论文 arXiv:2511.05403",
"url": "https://arxiv.org/abs/2511.05403"
},
{
"label": "Voxel51 子集",
"url": "https://docs.voxel51.com/dataset_zoo/datasets_hf/palm.html"
}
],
"team": []
},
{
"id": "egoexo4d",
"name": "Ego-Exo4D 手部姿态",
"year": "CVPR 2024",
"status": "p0",
"tags": [
"新增"
],
"source": "egoexo",
"method": "mvkp",
"gt": "21 个 3D 关键点",
"tier": "-",
"route": "R3",
"camera": "Aria RGB（110° 视场，卷帘快门）+ 外视角相机",
"rig": {
"ego": true,
"fisheye": true,
"exo": 5
},
"scale": "约 6.8 万人工 2D、34 万人工 3D；自动 3D 约 2100 万",
"fills": "真实场景的宽视场第一视角，对应 HOT3D 上暴露的相机域问题。",
"risk": "没有 MANO，只能监督 pose。",
"links": [
{
"label": "论文 arXiv:2311.18259",
"url": "https://arxiv.org/abs/2311.18259"
}
],
"team": [],
"subjects": ""
},
{
"id": "gigahands",
"name": "GigaHands",
"year": "CVPR 2025",
"status": "p1",
"tags": [
"新增"
],
"source": "exo",
"method": "mvkp",
"gt": "MANO",
"tier": "S4",
"route": "R1",
"camera": "51 台 RGB，只有外视角",
"rig": {
"exo": 51
},
"subjects": "56 人",
"scale": "34 小时，1.83 亿帧，8.4 万条文本描述",
"fills": "双手动作的多样性。",
"risk": "YOLO + HaMeR + ViTPose 初始化后三角化，再用 EasyMoCap 拟合，可能继承 HaMeR 的偏差，建议先抽检。",
"links": [
{
"label": "项目主页",
"url": "https://ivl.cs.brown.edu/research/gigahands.html"
},
{
"label": "论文 arXiv:2412.04244",
"url": "https://arxiv.org/abs/2412.04244"
}
],
"team": []
},
{
"id": "oakink2",
"name": "OakInk2",
"year": "CVPR 2024",
"status": "p1",
"tags": [
"新增"
],
"source": "egoexo",
"method": "marker",
"gt": "SMPL-X → MANO",
"tier": "S2",
"route": "R2",
"camera": "1 个第一视角 + 3 个外视角 RGB，848×480",
"rig": {
"ego": true,
"exo": 3
},
"subjects": "9 人",
"scale": "627 个序列，401 万帧",
"fills": "第一视角加光学真值的双手复杂任务。",
"risk": "团队 10.8 只用了它的动作分布，图像未入清单。画面中可能看得到 marker，需检查。分辨率低。",
"links": [
{
"label": "项目主页",
"url": "https://oakink.net/v2"
},
{
"label": "论文 arXiv:2403.19417",
"url": "https://arxiv.org/abs/2403.19417"
}
],
"team": []
},
{
"id": "hint",
"name": "HInt",
"year": "CVPR 2024",
"status": "p1",
"tags": [
"新增"
],
"source": "wild",
"method": "manual2d",
"gt": "21 点 2D + 逐点遮挡标签",
"tier": "-",
"route": "R4",
"camera": "Ego4D、EPIC VISOR、Hands23、NewDays 图像",
"rig": {
"wild": true,
"ego": true
},
"fills": "野外第一视角的 2D 监督，以及按遮挡分层的评测。",
"risk": "Ego4D 帧需单独申请许可。",
"links": [
{
"label": "GitHub（MIT）",
"url": "https://github.com/ddshan/hint"
}
],
"team": [],
"subjects": "",
"scale": ""
},
{
"id": "blurhand",
"name": "BlurHand",
"year": "CVPR 2023",
"status": "p1",
"tags": [
"新增"
],
"source": "exo",
"method": "mvkp",
"gt": "首、中、尾三帧 3D 真值",
"tier": "S4",
"route": "R1",
"camera": "由 InterHand2.6M 插帧合成运动模糊",
"rig": {
"exo": 12
},
"scale": "训练 121,839 张，测试 34,057 张",
"fills": "运动模糊，对应 V7 台本里的快速动作。",
"links": [
{
"label": "GitHub",
"url": "https://github.com/JaehaKim97/BlurHand_RELEASE"
},
{
"label": "论文 arXiv:2303.15417",
"url": "https://arxiv.org/abs/2303.15417"
}
],
"team": [],
"risk": "",
"subjects": ""
},
{
"id": "dartset",
"name": "DARTset",
"year": "NeurIPS 2022",
"status": "p1",
"tags": [
"新增"
],
"source": "synth",
"method": "synth",
"gt": "MANO 扩展（DART）",
"tier": "SY",
"route": "R1",
"camera": "合成渲染",
"rig": {
"synth": true
},
"scale": "约 80 万张，50 种饰品，325 种皮肤纹理",
"fills": "外观多样性：戒指、手表、美甲、纹身。",
"links": [
{
"label": "NeurIPS 论文",
"url": "https://proceedings.neurips.cc/paper_files/paper/2022/hash/f06d5ebd4ff40b40dd97e30cee632123-Abstract-Datasets_and_Benchmarks.html"
}
],
"team": [],
"risk": "",
"subjects": ""
},
{
"id": "humbi",
"name": "HUMBI Hand",
"year": "CVPR 2020",
"status": "p2",
"tags": [
"新增"
],
"source": "exo",
"method": "mvkp",
"gt": "MANO",
"tier": "S4",
"route": "R1",
"camera": "手部 68 个视角",
"rig": {
"exo": 68
},
"subjects": "全集 772 人，性别均衡，多种肤色",
"fills": "shape 分布广。",
"risk": "每人 shape 取关键点拟合结果的时间中值，只能当弱监督。",
"links": [
{
"label": "项目主页",
"url": "http://humbi-data.net"
},
{
"label": "论文 arXiv:2110.00119",
"url": "https://arxiv.org/abs/2110.00119"
}
],
"team": [],
"scale": ""
},
{
"id": "ourcapture",
"name": "单人手部采集（V6 / V7）",
"year": "2026",
"status": "own",
"tags": [
"自采"
],
"source": "egoexo",
"method": "marker",
"gt": "SOMA（目标）",
"tier": "S2",
"route": "R0",
"camera": "Femto Bolt 头戴 RGB（带 marker）+ 光学动捕",
"rig": {
"ego": true,
"exo": 6
},
"subjects": "3 人（计划增加）",
"scale": "V6 152 条 + V7 新增 51 条",
"fills": "极端抓取、手内换抓、手手接触、注视与速度变化。",
"risk": "手上 marker 出现在 RGB 中，需按已知位置做 inpainting。",
"links": [],
"team": []
},
{
"id": "assemblyhands",
"name": "AssemblyHands",
"year": "CVPR 2023",
"status": "no",
"tags": [
"不建议"
],
"source": "norgb",
"method": "mvkp",
"gt": "21 个 3D 关键点",
"tier": "-",
"route": "R3",
"camera": "第一视角为黑白鱼眼；外视角为 RGB",
"rig": {
"ego": true,
"fisheye": true,
"mono": true,
"exo": 8
},
"subjects": "34 人",
"scale": "约 300 万张，其中第一视角约 49 万张",
"fills": "只有外视角 RGB 可用。",
"risk": "第一视角不是真实 RGB，与排除 SHOW3D 的理由相同。",
"links": [
{
"label": "论文 arXiv:2304.12301",
"url": "https://arxiv.org/abs/2304.12301"
}
],
"team": []
},
{
"id": "show3d",
"name": "SHOW3D",
"year": "2026",
"status": "no",
"tags": [
"不建议",
"第一人称图册"
],
"source": "norgb",
"method": "mvkp",
"gt": "UmeTrack（另提供 MANO / MHR 转换）",
"tier": "S4",
"route": "R1",
"camera": "黑白鱼眼",
"rig": {
"ego": true,
"fisheye": true,
"mono": true
},
"fills": "团队已排除。",
"risk": "无标记多视角自动拟合，人工核验子集不具代表性；黑白鱼眼不满足真实 RGB 条件。",
"team": [
{
"label": "团队图册：第一人称具身数据集",
"url": "https://yangzijing-nine-source-hand-data-browser.static.hf.space/egocentric_embodied_datasets_20261020/index.html"
}
],
"links": [],
"subjects": "",
"scale": ""
},
{
"id": "hiphi",
"name": "HiPHI",
"year": "2026",
"status": "no",
"tags": [
"不建议"
],
"source": "norgb",
"method": "marker",
"gt": "BVH 骨架",
"tier": "-",
"route": "R3",
"camera": "无 RGB",
"rig": {
"none": true
},
"subjects": "132 人",
"scale": "617.5 小时（含镜像），约 2 亿帧",
"fills": "大规模全身与物体交互动捕，可作动作先验。",
"risk": "没有图像，手部细节未说明。",
"links": [
{
"label": "论文 arXiv:2608.16222",
"url": "https://arxiv.org/abs/2608.16222"
}
],
"team": []
},
{
"id": "eventegohands",
"name": "EventEgoHands++（EEH-R）",
"year": "2026",
"status": "no",
"tags": [
"不建议"
],
"source": "norgb",
"method": "device",
"gt": "MANO（手套关节拟合）",
"tier": "S4",
"route": "R1",
"camera": "DAVIS346 事件相机",
"rig": {
"ego": true,
"none": true
},
"subjects": "8 人",
"scale": "1,019,716 帧，约 2.36 小时",
"fills": "暗光与事件相机。",
"risk": "没有 RGB，公开情况未说明。",
"links": [
{
"label": "alphaXiv",
"url": "https://www.alphaxiv.org/abs/2609.17189"
}
],
"team": []
}
]
} |