Whether Lingo dataset and model include object motion?

[YudeZou]

Hello, I greatly appreciate your work, which is very impressive. While reading the paper and the code, I have several doubts.

1. Whether Lingo dataset includes object motion? The explanation of the TRUMANS dataset includes Object_pose and object_mat.npy. However, I did not see these in the Lingo dataset's explanation, may I ask if the Lingo dataset includes the motion of object?

2. Another question is, does the Lingo model generate object motion, too? In the paper, I see that it does, and may I ask which part of the code is related to this?

[hatw95]

I’m wondering the same thing.
In the demo and the paper, it seems that all objects are segmented and small objects can be manipulated.
However, in the dataset you provided, the scene mesh only contains a merged low-poly version.
Do you have any plans to release the individual object meshes?

[jiewjwang]

Hello, I greatly appreciate your work, which is very impressive. While reading the paper and the code, I have several doubts.

1. Whether Lingo dataset includes object motion? The explanation of the TRUMANS dataset includes Object_pose and object_mat.npy. However, I did not see these in the Lingo dataset's explanation, may I ask if the Lingo dataset includes the motion of object?
2. Another question is, does the Lingo model generate object motion, too? In the paper, I see that it does, and may I ask which part of the code is related to this?

I'd like to ask if you've solved this problem about interacting with objects? I look forward to your reply.

[Pcc123456]

Hello, I greatly appreciate your work, which is very impressive. While reading the paper and the code, I have several doubts.

1. Whether Lingo dataset includes object motion? The explanation of the TRUMANS dataset includes Object_pose and object_mat.npy. However, I did not see these in the Lingo dataset's explanation, may I ask if the Lingo dataset includes the motion of object?
2. Another question is, does the Lingo model generate object motion, too? In the paper, I see that it does, and may I ask which part of the code is related to this?

I'd like to ask if you've solved this problem about interacting with objects? I look forward to your reply.

Are you saying that the human body can grasp and move objects? It seems that I have also encountered this problem. The "vis.blender" provided by the author seems to be the main body of an entire scene. There is no content that can be moved, and I don't know why this scene doesn't have rendered colors.

[jiewjwang]

Hello, I greatly appreciate your work, which is very impressive. While reading the paper and the code, I have several doubts.

1. Whether Lingo dataset includes object motion? The explanation of the TRUMANS dataset includes Object_pose and object_mat.npy. However, I did not see these in the Lingo dataset's explanation, may I ask if the Lingo dataset includes the motion of object?
2. Another question is, does the Lingo model generate object motion, too? In the paper, I see that it does, and may I ask which part of the code is related to this?

I'd like to ask if you've solved this problem about interacting with objects? I look forward to your reply.

Are you saying that the human body can grasp and move objects? It seems that I have also encountered this problem. The "vis.blender" provided by the author seems to be the main body of an entire scene. There is no content that can be moved, and I don't know why this scene doesn't have rendered colors.

Yes, I've also encountered this problem. I saw that the latest method, HOSIG, renders the method from the Lingo paper, and the rendered scene is in color. I'm not sure how they achieved this rendering. I ​​also have another question: do you know how to change scenes in vis.blend? I'm not sure how to do it. I look forward to your reply.

[Pcc123456]

Hello, I greatly appreciate your work, which is very impressive. While reading the paper and the code, I have several doubts.

1. Whether Lingo dataset includes object motion? The explanation of the TRUMANS dataset includes Object_pose and object_mat.npy. However, I did not see these in the Lingo dataset's explanation, may I ask if the Lingo dataset includes the motion of object?
2. Another question is, does the Lingo model generate object motion, too? In the paper, I see that it does, and may I ask which part of the code is related to this?

I'd like to ask if you've solved this problem about interacting with objects? I look forward to your reply.

Are you saying that the human body can grasp and move objects? It seems that I have also encountered this problem. The "vis.blender" provided by the author seems to be the main body of an entire scene. There is no content that can be moved, and I don't know why this scene doesn't have rendered colors.

Yes, I've also encountered this problem. I saw that the latest method, HOSIG, renders the method from the Lingo paper, and the rendered scene is in color. I'm not sure how they achieved this rendering. I ​​also have another question: do you know how to change scenes in vis.blend? I'm not sure how to do it. I look forward to your reply.

Hello, I'm also just starting to learn this field. However, I would like to add some contact information such as WeChat so that we can study together. This is my email address: pccshuai@gmail.com