{
"cells": [
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Sat Mar 8 20:52:02 2025 \n",
"+-----------------------------------------------------------------------------+\n",
"| NVIDIA-SMI 525.116.04 Driver Version: 525.116.04 CUDA Version: 12.0 |\n",
"|-------------------------------+----------------------+----------------------+\n",
"| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n",
"| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n",
"| | | MIG M. |\n",
"|===============================+======================+======================|\n",
"| 0 NVIDIA GeForce ... On | 00000000:3D:00.0 Off | N/A |\n",
"| 30% 25C P8 5W / 250W | 6209MiB / 11264MiB | 0% Default |\n",
"| | | N/A |\n",
"+-------------------------------+----------------------+----------------------+\n",
" \n",
"+-----------------------------------------------------------------------------+\n",
"| Processes: |\n",
"| GPU GI CI PID Type Process name GPU Memory |\n",
"| ID ID Usage |\n",
"|=============================================================================|\n",
"+-----------------------------------------------------------------------------+\n"
]
}
],
"source": [
"!nvidia-smi"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"only got 89/128 samples!\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"only got 122/128 samples!\n",
"only got 124/128 samples!\n",
"only got 123/128 samples!\n",
"only got 125/128 samples!\n",
"only got 127/128 samples!\n",
"only got 121/128 samples!\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Processing link #0: base_link\n",
"Processing link #1: link_0.0\n",
"Processing link #2: link_1.0\n",
"Processing link #3: link_2.0\n",
"Processing link #4: link_3.0\n",
"Processing link #5: link_3.0_tip\n",
"Processing link #6: link_4.0\n",
"Processing link #7: link_5.0\n",
"Processing link #8: link_6.0\n",
"Processing link #9: link_7.0\n",
"Processing link #10: link_7.0_tip\n",
"Processing link #11: link_8.0\n",
"Processing link #12: link_9.0\n",
"Processing link #13: link_10.0\n",
"Processing link #14: link_11.0\n",
"Processing link #15: link_11.0_tip\n",
"Processing link #16: link_12.0\n",
"Processing link #17: link_13.0\n",
"Processing link #18: link_14.0\n",
"Processing link #19: link_15.0\n",
"Processing link #20: link_15.0_tip\n",
"Processing link #0: palm\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"only got 58/64 samples!\n",
"only got 63/64 samples!\n",
"only got 62/64 samples!\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Processing link #1: ffknuckle\n",
"Processing link #2: ffproximal\n",
"Processing link #3: ffmiddle\n",
"Processing link #4: ffdistal\n",
"Processing link #5: fftip\n",
"Processing link #6: mfknuckle\n",
"Processing link #7: mfproximal\n",
"Processing link #8: mfmiddle\n",
"Processing link #9: mfdistal\n",
"Processing link #10: mftip\n",
"Processing link #11: rfknuckle\n",
"Processing link #12: rfproximal\n",
"Processing link #13: rfmiddle\n",
"Processing link #14: rfdistal\n",
"Processing link #15: rftip\n",
"Processing link #16: lfmetacarpal\n",
"Processing link #17: lfknuckle\n",
"Processing link #18: lfproximal\n",
"Processing link #19: lfmiddle\n",
"Processing link #20: lfdistal\n",
"Processing link #21: lftip\n",
"Processing link #22: thbase\n",
"Processing link #23: thproximal\n",
"Processing link #24: thhub\n",
"Processing link #25: thmiddle\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"only got 126/128 samples!\n",
"only got 115/128 samples!\n",
"only got 124/128 samples!\n",
"only got 125/128 samples!\n",
"only got 122/128 samples!\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Processing link #26: thdistal\n",
"Processing link #27: thtip\n",
"Processing link #0: base_link\n",
"Processing link #1: palm\n",
"Processing link #2: a_link1\n",
"Processing link #3: a_link2\n",
"Processing link #4: a_link3\n",
"Processing link #5: a_link1_joint\n",
"Processing link #6: a_link1_dh\n",
"Processing link #7: a_link2_joint\n",
"Processing link #8: a_link2_dh\n",
"Processing link #9: a_link3_joint\n",
"Processing link #10: a_link3_dh\n",
"Processing link #11: b_link1\n",
"Processing link #12: b_link2\n",
"Processing link #13: b_link3\n",
"Processing link #14: b_link1_joint\n",
"Processing link #15: b_link1_dh\n",
"Processing link #16: b_link2_joint\n",
"Processing link #17: b_link2_dh\n",
"Processing link #18: b_link3_joint\n",
"Processing link #19: b_link3_dh\n",
"Processing link #20: c_link2\n",
"Processing link #21: c_link3\n",
"Processing link #22: c_link2_joint\n",
"Processing link #23: c_link2_joint_0\n",
"Processing link #24: c_link2_dh\n",
"Processing link #25: c_link3_joint\n",
"Processing link #26: c_link3_dh\n",
"Processing link #0: base_link\n",
"Processing link #1: jaco_6_hand_limb\n",
"Processing link #2: jaco_fingers_base_link\n",
"Processing link #3: jaco_7_finger_mount_index\n",
"Processing link #4: jaco_8_finger_index\n",
"Processing link #5: jaco_9_finger_index_tip\n",
"Processing link #6: jaco_7_finger_mount_thumb\n",
"Processing link #7: jaco_8_finger_thumb\n",
"Processing link #8: jaco_9_finger_thumb_tip\n",
"Processing link #9: jaco_7_finger_mount_pinkie\n",
"Processing link #10: jaco_8_finger_pinkie\n",
"Processing link #11: jaco_9_finger_pinkie_tip\n"
]
}
],
"source": [
"ROOT_DIR = \"/data/GraspLLM_open/Multi-GraspSet\"\n",
"import os\n",
"import sys\n",
"sys.path.append(ROOT_DIR)\n",
"os.chdir(ROOT_DIR)\n",
"\n",
"\n",
"from utils.hand_model import HandModel\n",
"from utils.panda_gripper import PandaGripper\n",
"import json\n",
"\n",
"\n",
"urdf_allegro_path=\"./hand_model/allegro_hand_description/allegro_hand_description_right.urdf\"\n",
"meshes_allegro_path=\"./hand_model/allegro_hand_description/meshes\"\n",
"allegro_hand_model = HandModel(\"allegro\", urdf_allegro_path, meshes_allegro_path, batch_size=1, device=\"cuda\", hand_scale=1)\n",
"\n",
"urdf_shadow_path=\"./hand_model/shadow_hand_description/urdf/shadowhand_adjust.urdf\"\n",
"meshes_shadow_path=\"./hand_model/shadow_hand_description/meshes\"\n",
"shadow_hand_model = HandModel(\"shadowhand\", urdf_shadow_path, meshes_shadow_path, batch_size=1, device=\"cuda\", hand_scale=1)\n",
"\n",
"\n",
"panda_gripper=PandaGripper()\n",
"\n",
"\n",
"urdf_barrett_path=\"./hand_model/barrett_hand_description/Barrett.urdf\"\n",
"meshes_barrett_path=\"./hand_model/barrett_hand_description/meshes\"\n",
"barrett_hand_model = HandModel(\"barrett\", urdf_barrett_path, meshes_barrett_path, batch_size=1, device=\"cuda\", hand_scale=1)\n",
"\n",
"urdf_jaco_path=\"./hand_model/jaco_hand_description/jaco_robot.urdf\"\n",
"meshes_jaco_path=\"./hand_model/jaco_hand_description/\"\n",
"jaco_hand_model = HandModel(\"jaco\", urdf_jaco_path, meshes_jaco_path, batch_size=1, device=\"cuda\", hand_scale=1)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"import pickle\n",
"\n",
"with open(\"./mid-level-data/shadow.pkl\",\"rb\") as f:\n",
" meta_shadow=pickle.load(f)\n",
"\n",
"with open(\"./mid-level-data/allegro.pkl\",\"rb\") as f:\n",
" meta_allegro=pickle.load(f)\n",
" \n",
"with open(\"./mid-level-data/panda.pkl\",\"rb\") as f:\n",
" meta_panda=pickle.load(f)\n",
" \n",
"with open(\"./mid-level-data/barrett.pkl\",\"rb\") as f:\n",
" meta_barret=pickle.load(f)\n",
" \n",
"with open(\"./mid-level-data/jaco.pkl\",\"rb\") as f:\n",
" meta_jaco=pickle.load(f) \n",
" "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualize Allegro Hand"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Middle finger,thumb finger and index finger touch the bottle body of cylinder_bottle,little finger touches the bottle cap of cylinder_bottle.\n"
]
},
{
"data": {
"text/html": [
""
],
"text/plain": [
""
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import trimesh\n",
"import re\n",
"import torch\n",
"import numpy as np\n",
"vocab_size=511\n",
"object_code=\"cylinder_bottle_s104\"\n",
"num=0\n",
"\n",
"angle=meta_allegro[object_code][num]['angle']\n",
"obj_id=object_code\n",
"\n",
"\n",
"infer_pose=torch.tensor(angle[0]+angle[1][:, :2].T.ravel().tolist()+angle[2])\n",
"mesh_hand_list_infer=allegro_hand_model.get_meshes_from_q(q=infer_pose.cuda().unsqueeze(0))\n",
"mesh_hand= trimesh.util.concatenate(mesh_hand_list_infer)\n",
"\n",
"pc=np.load(f\"./pc_8192/{obj_id}.npy\")[:,:3]\n",
"point_colors = np.tile([255, 0, 0, 255], (pc.shape[0], 1))\n",
"point_cloud = trimesh.points.PointCloud(pc,colors=point_colors)\n",
"mesh_obj=trimesh.load(f\"./Mesh_nor/{object_code}.obj\",force=\"mesh\")\n",
"\n",
"print(meta_allegro[object_code][num]['contact_info'])\n",
"scene = trimesh.Scene()\n",
"scene.add_geometry(point_cloud)\n",
"# scene.add_geometry(mesh_obj)\n",
"scene.add_geometry(mesh_hand)\n",
"\n",
"scene.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualize Shadow Hand"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"['Little finger,ring finger,middle finger,thumb finger and index finger touch the mug body of mug.', 'all fingers touches the mug body of mug']\n"
]
},
{
"data": {
"text/html": [
""
],
"text/plain": [
""
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import trimesh\n",
"import re\n",
"vocab_size=511\n",
"object_code=\"mug_s139\"\n",
"num=1\n",
"angle=meta_shadow[object_code][num]['angle']\n",
"obj_id=object_code\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"infer_pose=torch.tensor(angle[0]+angle[1][:, :2].T.ravel().tolist()+angle[2])\n",
"mesh_hand_list_infer=shadow_hand_model.get_meshes_from_q(q=infer_pose.cuda().unsqueeze(0))\n",
"mesh_hand= trimesh.util.concatenate(mesh_hand_list_infer)\n",
"\n",
"print(meta_shadow[object_code][num]['contact_info'])\n",
"pc=np.load(f\"./pc_8192/{obj_id}.npy\")[:,:3]\n",
"point_colors = np.tile([255, 0, 0, 255], (pc.shape[0], 1))\n",
"point_cloud = trimesh.points.PointCloud(pc,colors=point_colors)\n",
"scene = trimesh.Scene()\n",
"scene.add_geometry(point_cloud)\n",
"scene.add_geometry(mesh_hand)\n",
"scene.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visual Panda Gripper"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"dict_keys(['pump head', 'bottle body'])\n"
]
},
{
"data": {
"text/html": [
""
],
"text/plain": [
""
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"object_code=\"chaoneng_handwash\"\n",
"print(meta_panda[object_code].keys())\n",
"\n",
"angle=meta_panda[object_code]['pump head'][0]\n",
"mesh_hand=panda_gripper.get_meshes(angle)\n",
"\n",
"pc=np.load(f\"./pc_8192/{object_code}.npy\")[:,:3]\n",
"point_colors = np.tile([255, 0, 0, 255], (pc.shape[0], 1))\n",
"point_cloud = trimesh.points.PointCloud(pc,colors=point_colors)\n",
"scene = trimesh.Scene()\n",
"scene.add_geometry(point_cloud)\n",
"scene.add_geometry(mesh_hand)\n",
"scene.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Visualize Barrett"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"([-0.014817648685191163, 0.115100027408388, -0.17385810370641025],\n",
" array([[-0.0887277 , 0.90763646, 0.41027241],\n",
" [ 0.19694446, 0.41975144, -0.88601445],\n",
" [-0.97639146, 0.00218686, -0.21599755]]),\n",
" [0.820802, 0.813802, 1.05])"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"angle"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"all fingers touches the mug body of mug\n"
]
},
{
"data": {
"text/html": [
""
],
"text/plain": [
""
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import trimesh\n",
"import re\n",
"import torch\n",
"import numpy as np\n",
"vocab_size=511\n",
"object_code=\"mug_s139\"\n",
"num=0\n",
"angle=meta_barret[object_code][num]['angle']\n",
"obj_id=object_code\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"infer_pose=torch.tensor(angle[0]+angle[1][:, :2].T.ravel().tolist()+angle[2])\n",
"mesh_hand_list_infer=barrett_hand_model.get_meshes_from_q(q=infer_pose.cuda().unsqueeze(0))\n",
"mesh_hand= trimesh.util.concatenate(mesh_hand_list_infer)\n",
"\n",
"print(meta_barret[object_code][num]['contact_info'])\n",
"pc=np.load(f\"./pc_8192/{object_code}.npy\")[:,:3]\n",
"point_colors = np.tile([255, 0, 0, 255], (pc.shape[0], 1))\n",
"point_cloud = trimesh.points.PointCloud(pc,colors=point_colors)\n",
"scene = trimesh.Scene()\n",
"scene.add_geometry(point_cloud)\n",
"scene.add_geometry(mesh_hand)\n",
"scene.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Jaco"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"2 fingers touches the mug handle of mug\n"
]
},
{
"data": {
"text/html": [
""
],
"text/plain": [
""
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import trimesh\n",
"import re\n",
"import torch\n",
"import numpy as np\n",
"vocab_size=511\n",
"object_code=\"mug_s139\"\n",
"num=0\n",
"angle=meta_jaco[object_code][num]['angle']\n",
"obj_id=object_code\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"infer_pose=torch.tensor(angle[0]+angle[1][:, :2].T.ravel().tolist()+angle[2])\n",
"mesh_hand_list_infer=jaco_hand_model.get_meshes_from_q(q=infer_pose.cuda().unsqueeze(0))\n",
"mesh_hand= trimesh.util.concatenate(mesh_hand_list_infer)\n",
"\n",
"print(meta_jaco[object_code][num]['contact_info'])\n",
"pc=np.load(f\"./pc_8192/{object_code}.npy\")[:,:3]\n",
"point_colors = np.tile([255, 0, 0, 255], (pc.shape[0], 1))\n",
"point_cloud = trimesh.points.PointCloud(pc,colors=point_colors)\n",
"scene = trimesh.Scene()\n",
"scene.add_geometry(point_cloud)\n",
"scene.add_geometry(mesh_hand)\n",
"scene.show()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [conda env:eval]",
"language": "python",
"name": "conda-env-eval-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.19"
}
},
"nbformat": 4,
"nbformat_minor": 2
}