image listlengths 2 8 | conversations listlengths 2 2 |
|---|---|
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_device_100013_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_device_100013_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_device_100013_view5.png",
"datasets/my-data... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100015_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100015_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100015_view5.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100017_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100017_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100017_view7.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100021_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100021_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100021_view3.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100023_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100023_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100023_view2.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100025_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100025_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100032_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100032_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100033_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100033_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100033_view8.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100038_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100038_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100038_view7.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100040_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100040_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100040_view2.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100045_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100045_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100045_view3.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100047_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100047_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100051_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100051_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100051_view9.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100054_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100054_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100054_view8.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100056_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100056_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100056_view3.png",
"datasets/my-dataset/physx_image_aug2/cookware/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100057_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100057_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100060_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100060_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/cookware/physx_cookware_100060_view2.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100061_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100061_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100061_view4.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100064_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100064_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100064_view8.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100068_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100068_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100068_view4.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100071_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100071_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100071_view2.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100072_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100072_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100073_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100073_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100073_view7.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/transport/physx_transport_equipment_100075_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/transport/physx_transport_equipment_100075_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/transport/physx_transport_equipment_100075_view7.png",
"datasets/my-d... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100079_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100079_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100079_view4.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100082_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100082_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100082_view4.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100085_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100085_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100085_view1.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100086_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100086_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100087_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100087_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100092_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100092_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100092_view9.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100095_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100095_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100095_view9.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100103_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100103_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100103_view6.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100108_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100108_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100109_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100109_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100109_view6.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100116_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100116_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100116_view3.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100123_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100123_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100128_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100128_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100128_view8.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100133_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100133_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/electronic/physx_electronic_storage_device_100133_view5.p... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100142_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100142_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100142_view9.png",
"datasets/my-dataset/physx_image_aug2/handtool/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100144_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100144_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100144_view5.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100146_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100146_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100146_view7.png",
"datasets/my-dataset/physx_image_aug2/handtool/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100150_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100150_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/handtool/physx_handtool_100150_view4.png",
"datasets/my-dataset/physx_image_aug2/handtool/phy... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100154_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100154_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100154_view1.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100162_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100162_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100162_view9.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100174_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100174_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100174_view2.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100179_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100179_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100179_view5.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100180_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100180_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100180_view2.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100182_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100182_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_100182_view5.png",
"datasets/my-dataset/physx_image_aug2/hand/physx_hand_tool_... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100189_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100189_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100189_view8.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/packaging/physx_packaging_container_100191_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/packaging/physx_packaging_container_100191_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/packaging/physx_packaging_container_100191_view5.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100194_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100194_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100194_view2.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100197_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100197_canonical_seg.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 2 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendered in a un... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100202_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100202_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100202_view3.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100214_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100214_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100214_view7.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/packaging/physx_packaging_container_100224_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/packaging/physx_packaging_container_100224_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/packaging/physx_packaging_container_100224_view3.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100234_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100234_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100234_view2.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100243_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100243_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100243_view8.png"
] | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 3 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct region rendere... |
[
"datasets/my-dataset/physx_image_aug2/container/physx_container_100247_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100247_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/container/physx_container_100247_view3.png",
"datasets/my-dataset/physx_image_aug2/contai... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
[
"datasets/my-dataset/physx_image_aug2/luggage/physx_luggage_100248_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/luggage/physx_luggage_100248_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/luggage/physx_luggage_100248_view6.png",
"datasets/my-dataset/physx_image_aug2/luggage/physx_lugg... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 4 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each distinct regio... |
[
"datasets/my-dataset/physx_image_aug2/luggage/physx_luggage_100249_canonical_view.png",
"datasets/my-dataset/physx_image_aug2/luggage/physx_luggage_100249_canonical_seg.png",
"datasets/my-dataset/physx_image_aug2/luggage/physx_luggage_100249_view2.png",
"datasets/my-dataset/physx_image_aug2/luggage/physx_lugg... | [
{
"from": "human",
"value": "<image>\n<image>\n<image>\n<image>\n<image>\n\nYou are a 3D reconstruction agent.\n\nYou are given 5 images of an object/scene.\n - Image 1: rendered view of the object/scene with full materials.\n - Image 2: segmentation map at the SAME camera as Image 1, with\n each disti... |
End of preview. Expand in Data Studio
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
5-Mix Training Dataset for Qwen-VL 3DCoder
The dataset used to train sft_4b_canon_aug2_v13mv_v13aug2_scenev4_mv —
a Qwen3-VL-4B fine-tune that predicts part-level kinematic structure
(materials, parent tree, joints, relations) from a single rendered view +
its colour segmentation map.
Contents
jsonl/ structure-prediction JSONLs
physx_canon_structure_train.jsonl 4 895 PhysX-Anything (canonical Z-up frame)
physx_canon_aug2_structure_train.jsonl 4 899 with split/merge augmentation
v13_njoint_root_zup_mv_structure_train.jsonl 6 104 single-object v13 multi-view
v13_njoint_root_zup_aug2_structure_train.jsonl 4 748 with split/merge aug
scene_v4_unified_root_v2_train.jsonl 1 730 multi-object scene_v4
physx_canon_structure_test.jsonl 387 PhysX-Anything test split
v13_njoint_root_zup_mv_structure_test.jsonl 678 v13 test split
scene_v4_unified_root_v2_test.jsonl 192 scene_v4 test split
physx_image/ PhysX canonical-view renders + segmaps
scene_image/ multi-object scene_v4 renders + segmaps
v13_image/ v13 single-object multi-view renders + segmaps
sft_qwen3_4b_canon_aug2_v13mv_v13aug2_scenev4_mv.sh the exact training launch script
JSONL schema
Each row is a conversation pair (image-conditioned text-to-JSON):
{
"image": ["physx_image/...rendered.png", "physx_image/...segmap.png", ...],
"conversations": [
{"from": "human", "value": "<image>\n<image>\n\nYou are a 3D reconstruction agent.\n..."},
{"from": "gpt", "value": "{\"materials\":{...},\"nodes\":{...},\"relations\":[...]}"}
]
}
All centroids/extents and joint positions are in canonical Z-up,
unit-cube-normalised coordinates (combined-bbox longest axis = 1.0).
The frame convention: +X = front, +Z = up, ±Y = lateral.
Coordinate frame
| family | source frame | rotation to canonical |
|---|---|---|
physx_canon_* |
PhysX raw (+Y=front, -Z=up) |
(x,y,z) → (z,x,y) |
v13_* |
infinigen raw (+Y=up) |
R_x(π/2) |
scene_v4_* |
scene raw | per-scene-graph normalisation |
The training launch script is preserved at the repo root so the exact
--dataset_use recipe and hyperparameters can be inspected.
License
MIT (data derived from PhysX-Anything and Infinigen procedural renders).
- Downloads last month
- 13,046