Thank you to everyone who participated in the Physically Informed 3D Food Reconstruction competition. We are now entering Phase 2 of the competition, where we will select the top 5 teams to submit their 3D object models. These models will be evaluated based on chamfer distance. The final score will be calculated by averaging the rankings from Phase 1 and Phase 2. In the event of a tie, the Phase 1 score will determine the final ranking.
\nFor the Phase 2 evaluation (for the top 5 teams), we were unable to find a fair and automatic way to align the pose between the submissions and the ground truth objects. Therefore, we are updating the evaluation procedure, and the winner announcement will be delayed to June 7th. The revised steps are as follows:
\nThank you for your understanding and cooperation.
", "messages": [ { "id": "2836515", "postDate": "05/26/2024 01:25:08", "content": "Thank you to everyone who participated in the Physically Informed 3D Food Reconstruction competition. We are now entering Phase 2 of the competition, where we will select the top 5 teams to submit their 3D object models. These models will be evaluated based on chamfer distance. The final score will be calculated by averaging the rankings from Phase 1 and Phase 2. In the event of a tie, the Phase 1 score will determine the final ranking.
\nFor the Phase 2 evaluation (for the top 5 teams), we were unable to find a fair and automatic way to align the pose between the submissions and the ground truth objects. Therefore, we are updating the evaluation procedure, and the winner announcement will be delayed to June 7th. The revised steps are as follows:
\nThank you for your understanding and cooperation.
", "rawMarkdown": "Thank you to everyone who participated in the Physically Informed 3D Food Reconstruction competition. We are now entering Phase 2 of the competition, where we will select the top 5 teams to submit their 3D object models. These models will be evaluated based on chamfer distance. The final score will be calculated by averaging the rankings from Phase 1 and Phase 2. In the event of a tie, the Phase 1 score will determine the final ranking.\n\nFor the Phase 2 evaluation (for the top 5 teams), we were unable to find a fair and automatic way to align the pose between the submissions and the ground truth objects. Therefore, we are updating the evaluation procedure, and the winner announcement will be delayed to June 7th. The revised steps are as follows:\n\n1. Upload Models: Upload all 20 of your 3D models (.obj files named with the object ID) to a location where we can download them.\n2. Send Link: Send us the link to your models along with your team ID (or other ways to identify yourself) by May 27th, 11:59 PM AOE. Email: metafood.cvpr@gmail.com. If you encounter any issues, please post on the Kaggle discussion forum.\n3. Model Verification: We will check your models against your final Phase 1 submission to ensure they are the same models used in Phase 1. We will also visually inspect the models to ensure no team is violating the rules by submitting primitive objects such as spheres.\n4. Ground Truth Models: We will upload our ground truth 3D models, along with the script used to compute the final chamfer distance, on Kaggle by May 28th, 11:59 PM AOE.\n5. Align Models: Align the models yourself and prepare a transform matrix for each object you submitted. Your final chamfer distance score will be computed using your submitted models and transform matrices.\n6. Submit Transform Matrices: Send us your transform matrices by June 4th, 11:59 PM AOE.\nWe will announce the winner on June 7th.\n\nThank you for your understanding and cooperation.", "votes": null }, { "id": "2839568", "postDate": "05/27/2024 15:28:04", "content": "Thank you for your efforts in the competition. I would like to know if the Chamfer distance in the phase 2 is calculated using the sum or the mean?
", "rawMarkdown": "Thank you for your efforts in the competition. I would like to know if the Chamfer distance in the phase 2 is calculated using the sum or the mean?", "votes": null }, { "id": "2839713", "postDate": "05/27/2024 16:54:18", "content": "One more question~ Are the meshes of the 20 objects waterproof, or is the side that is close to the the table missing?
", "rawMarkdown": "One more question~ Are the meshes of the 20 objects waterproof, or is the side that is close to the the table missing?", "votes": null }, { "id": "2839789", "postDate": "05/27/2024 17:47:37", "content": "They are waterproof. They are scanned by a scanner.
", "rawMarkdown": "They are waterproof. They are scanned by a scanner.", "votes": null }, { "id": "2839845", "postDate": "05/27/2024 18:19:44", "content": "Thank you so much for organizing the challenge. We, the VolETA team, tried to submit our results (the 20 3D models) using Kaggle, but unfortunately, we can only upload CSV files. Instead, we send an email to metafood.cvpr@gmail.com. Can you please confirm via email that you received the data? looking forward to hearing from you soon!
", "rawMarkdown": "Thank you so much for organizing the challenge. We, the VolETA team, tried to submit our results (the 20 3D models) using Kaggle, but unfortunately, we can only upload CSV files. Instead, we send an email toMean will be used when calculating the Chamfer distance for each direction. Final score for a model is the sum of the two chamfer distance under two directions. Reference: (https://pytorch3d.readthedocs.io/en/latest/_modules/pytorch3d/loss/chamfer.html)
", "rawMarkdown": "Mean will be used when calculating the Chamfer distance for each direction. Final score for a model is the sum of the two chamfer distance under two directions. Reference: (https://pytorch3d.readthedocs.io/en/latest/_modules/pytorch3d/loss/chamfer.html)", "votes": null }, { "id": "2839947", "postDate": "05/27/2024 19:58:36", "content": "We received your submission. Thank you!
", "rawMarkdown": "We received your submission. Thank you!", "votes": null } ], "comments": [ { "id": 2839568, "author_name": "tangyi32", "author_url": "", "post_date": "05/27/2024 15:28:04", "content": "Thank you for your efforts in the competition. I would like to know if the Chamfer distance in the phase 2 is calculated using the sum or the mean?
", "votes": null, "replies": [ { "id": 2839944, "author_name": "metafoodcvpr", "author_url": "", "post_date": "05/27/2024 19:53:11", "content": "Mean will be used when calculating the Chamfer distance for each direction. Final score for a model is the sum of the two chamfer distance under two directions. Reference: (https://pytorch3d.readthedocs.io/en/latest/_modules/pytorch3d/loss/chamfer.html)
", "votes": null, "replies": [] } ] }, { "id": 2839713, "author_name": "tangyi32", "author_url": "", "post_date": "05/27/2024 16:54:18", "content": "One more question~ Are the meshes of the 20 objects waterproof, or is the side that is close to the the table missing?
", "votes": null, "replies": [ { "id": 2839789, "author_name": "metafoodcvpr", "author_url": "", "post_date": "05/27/2024 17:47:37", "content": "They are waterproof. They are scanned by a scanner.
", "votes": null, "replies": [] } ] }, { "id": 2839845, "author_name": "umairharoon", "author_url": "", "post_date": "05/27/2024 18:19:44", "content": "Thank you so much for organizing the challenge. We, the VolETA team, tried to submit our results (the 20 3D models) using Kaggle, but unfortunately, we can only upload CSV files. Instead, we send an email to metafood.cvpr@gmail.com. Can you please confirm via email that you received the data? looking forward to hearing from you soon!
", "votes": null, "replies": [ { "id": 2839947, "author_name": "metafoodcvpr", "author_url": "", "post_date": "05/27/2024 19:58:36", "content": "We received your submission. Thank you!
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2836515": "Thank you to everyone who participated in the Physically Informed 3D Food Reconstruction competition. We are now entering Phase 2 of the competition, where we will select the top 5 teams to submit their 3D object models. These models will be evaluated based on chamfer distance. The final score will be calculated by averaging the rankings from Phase 1 and Phase 2. In the event of a tie, the Phase 1 score will determine the final ranking.\n\nFor the Phase 2 evaluation (for the top 5 teams), we were unable to find a fair and automatic way to align the pose between the submissions and the ground truth objects. Therefore, we are updating the evaluation procedure, and the winner announcement will be delayed to June 7th. The revised steps are as follows:\n\n1. Upload Models: Upload all 20 of your 3D models (.obj files named with the object ID) to a location where we can download them.\n2. Send Link: Send us the link to your models along with your team ID (or other ways to identify yourself) by May 27th, 11:59 PM AOE. Email: metafood.cvpr@gmail.com. If you encounter any issues, please post on the Kaggle discussion forum.\n3. Model Verification: We will check your models against your final Phase 1 submission to ensure they are the same models used in Phase 1. We will also visually inspect the models to ensure no team is violating the rules by submitting primitive objects such as spheres.\n4. Ground Truth Models: We will upload our ground truth 3D models, along with the script used to compute the final chamfer distance, on Kaggle by May 28th, 11:59 PM AOE.\n5. Align Models: Align the models yourself and prepare a transform matrix for each object you submitted. Your final chamfer distance score will be computed using your submitted models and transform matrices.\n6. Submit Transform Matrices: Send us your transform matrices by June 4th, 11:59 PM AOE.\nWe will announce the winner on June 7th.\n\nThank you for your understanding and cooperation.", "2839568": "Thank you for your efforts in the competition. I would like to know if the Chamfer distance in the phase 2 is calculated using the sum or the mean?", "2839713": "One more question~ Are the meshes of the 20 objects waterproof, or is the side that is close to the the table missing?", "2839789": "They are waterproof. They are scanned by a scanner.", "2839845": "Thank you so much for organizing the challenge. We, the VolETA team, tried to submit our results (the 20 3D models) using Kaggle, but unfortunately, we can only upload CSV files. Instead, we send an email to