Hello code mental, can you please give the code in GitHub and at 29:34 what is mean by label and where it is coming from. i am getting error as label is not defined. why?
thank you for your helpful. which algorithm is best to model 3D from 2d image using deep learning techniques and how can I train my own custom datasets !
Hello sir we are using the electronic applications like an power adapter or a 12v battery to be constructed in 3d model by 2d images using python or machine learning is it possible by this method if so then please share us the algorithm or if you know any of the method to reconstruction of the 2d image into 3d model then please convey us it will be very helpful
If you could somehow generate different perspectives of the object using a GAN, then you could use a differential renderer to reconstruct your object onto 3d. Actually that's how Nvidia has done it with GanVerse3D! Check it out, I have videos on the channel. But all of this is ground breaking research, still very early stages.
Videos you probably want to watch: ✅ 3D Deep Learning 👉🏼 th-cam.com/play/PL3OV2Akk7XpBPBEw1jekpxDJYIRPEHbUi.html ✅ Creating a 3D model of a Photo using Python and Numpy-STL 👉🏼 th-cam.com/play/PL3OV2Akk7XpDs8uSg6iegIKlmY8XxSrNr.html ✅ 3D Deep Learning tutorial with Nvidia Kaolin and PyTorch 👉🏼 th-cam.com/video/BwZ7s7DqbFs/w-d-xo.html ✅ 3D Deep Learning tutorial with Pytorch3d 👉 th-cam.com/video/v3hTD9m2tM8/w-d-xo.html ✅ Nvidia GANverse3D - 2D to 3D Model 👉🏼 th-cam.com/video/r5E4ThTukc4/w-d-xo.html ✅ 2D Photo to 3D Model by Google Phorhum 👉🏼 th-cam.com/video/rQoaJGvJYLc/w-d-xo.html ✅ Blender Python API 👉🏼 th-cam.com/play/PL3OV2Akk7XpBlVnaa6vqEWKcbg_sBTlPX.html ✅ Python Problems for Beginners 👉🏼 th-cam.com/play/PL3OV2Akk7XpC-fsHuJ3RLpnzOXmED7oAI.html ✅ How I created my 3D avatar with color from a 2D photo with PIFu 👉🏼 th-cam.com/video/o6dBgzX1pFE/w-d-xo.html
thank you for your helpful video. sir, which algorithm is best to model 3D from 2d image using deep learning techniques and how can I train my own custom datasets?
@@CODEMENTAL hi sir, i wanna implemented or involve fourier, or even fractional fourier to the 3d dataset, what is the first thing that i've should do?
Unlike ganverse3d you really need to know 3d deep learning to make good use of the tools provided by pytorch 3d. Check out pifuhd for something that you can try yourself. I have also made a video about it.
Hello sir can I apply this method or maybe you can share some algorithm?, which is use for reconstructing multiple 2D images from camera or 3 setup camera facing pile of flour bags into a 3D image/model, example of this 2d images is pile of flour bags because I want to reconstruct it into 3d and create counting system for how many bags in there. Thankyou sir I hope you can give your insight I'll subscribe to you
Not on its own. The differential renderer in the video is more of a tool that makes it possible to train machine learning models capable of reconstructing 3D objects using 2D training data. Photogrammetry is a technique to reconstruct 3D from multiple 2D images. Meshroom is an open-source software that you could try today, but I doubt it would provide good enough results for what you need. But do you really need to reconstruct 3D from 2D? What you are trying to do, sounds more like a 3D object detection problem. If I were you I would start with a paper like this one: arxiv.org/abs/2110.06922 Or, perhaps start simpler. You could also try to train an image classifier to count the number of flour bags based on on multiple images, without ever worrying about 3D. But I am just speculating. This sounds like a research project.
Stopping with your presentation grimaces in the foreground is ugly and counterproductive. For professional, serious and effective use, put a sober and simple presentation image.
a beautiful tutorial that god will bless you jeovanny
Very useful video thank you so much for your effort
Thank you
Thanks for your video. I would like to know how you jump directly to open the function definition in jupyter-lab.8:35
thanks I understand at least what needed to be done
hello brother where is the code brother?
Thank you for your presentation. It was very helpful.
Thank you for the feedback!
Hello code mental, can you please give the code in GitHub and at 29:34 what is mean by label and where it is coming from. i am getting error as label is not defined. why?
Instead of working with the vertex colour, can you work with the texture image
Would you mind giving any access to the code. I can't really tell what you've imported 😅
can you please give the google colab, i can't find it in the description, thank you!
Can I use this video to generate photo realistic 3D ?
hey if u dont mind can i get the code becuse i want to see a few things in that
liked ur presentation
I dont always model cows.. but when I do I only model half then merge it with the shadow.
hello thank you! does this process apply to custom images and obj files instead of the files found in google colab originally?
Hello, please help me out. I want to use 2d images and then create a 3d mesh with texture from those images. how to use this code for that?
This code is basically just to provide a framework for you to do do 3d deep learning. It doesn't actually give you the final solution for 3d to 2d
thank you for your helpful. which algorithm is best to model 3D from 2d image using deep learning techniques and how can I train my own custom datasets !
Pifu is a reference paper you should look into
Why not import a 2D image and transform it into a 3D model in a simple and fast way?
Wonderful!!
Hi Georgia, Really impressed by what the pytorch3d team has been up to! Looking forward to check what's new in the framework
so , would it be possible to make a tree using cropped tree pictures doing it this way?
You mean using differential rendering?
Hello sir we are using the electronic applications like an power adapter or a 12v battery to be constructed in 3d model by 2d images using python or machine learning is it possible by this method if so then please share us the algorithm or if you know any of the method to reconstruction of the 2d image into 3d model then please convey us it will be very helpful
Hey same here... Did u find anything
@@spatil689 i have converted 2d image in 3d image by using a python but we cant convert it into a 3d model.
If you find any convey me
Is this method applicable to a dataset of images in which we only have images with camera perspectives from the front of a given object?
If you could somehow generate different perspectives of the object using a GAN, then you could use a differential renderer to reconstruct your object onto 3d.
Actually that's how Nvidia has done it with GanVerse3D! Check it out, I have videos on the channel. But all of this is ground breaking research, still very early stages.
Videos you probably want to watch:
✅ 3D Deep Learning 👉🏼 th-cam.com/play/PL3OV2Akk7XpBPBEw1jekpxDJYIRPEHbUi.html
✅ Creating a 3D model of a Photo using Python and Numpy-STL 👉🏼 th-cam.com/play/PL3OV2Akk7XpDs8uSg6iegIKlmY8XxSrNr.html
✅ 3D Deep Learning tutorial with Nvidia Kaolin and PyTorch 👉🏼 th-cam.com/video/BwZ7s7DqbFs/w-d-xo.html
✅ 3D Deep Learning tutorial with Pytorch3d 👉 th-cam.com/video/v3hTD9m2tM8/w-d-xo.html
✅ Nvidia GANverse3D - 2D to 3D Model 👉🏼 th-cam.com/video/r5E4ThTukc4/w-d-xo.html
✅ 2D Photo to 3D Model by Google Phorhum 👉🏼 th-cam.com/video/rQoaJGvJYLc/w-d-xo.html
✅ Blender Python API 👉🏼 th-cam.com/play/PL3OV2Akk7XpBlVnaa6vqEWKcbg_sBTlPX.html
✅ Python Problems for Beginners 👉🏼 th-cam.com/play/PL3OV2Akk7XpC-fsHuJ3RLpnzOXmED7oAI.html
✅ How I created my 3D avatar with color from a 2D photo with PIFu 👉🏼 th-cam.com/video/o6dBgzX1pFE/w-d-xo.html
thank you for your helpful video. sir, which algorithm is best to model 3D from 2d image using deep learning techniques and how can I train my own custom datasets?
Sorry for late reply. Check Pifu. It has been used now in a few papers
@@CODEMENTAL hi sir, i wanna implemented or involve fourier, or even fractional fourier to the 3d dataset, what is the first thing that i've should do?
can you use your own image?
Unlike ganverse3d you really need to know 3d deep learning to make good use of the tools provided by pytorch 3d. Check out pifuhd for something that you can try yourself. I have also made a video about it.
@ ganverse3d works with objects but for now only with cars.
If you want to understand in more depth the research behind ganverse3d I recommend you to watch the video on Nvidia kaolin.
bruh thx for ur hard work XD
Thank you!
Hello sir can I apply this method or maybe you can share some algorithm?, which is use for reconstructing multiple 2D images from camera or 3 setup camera facing pile of flour bags into a 3D image/model, example of this 2d images is pile of flour bags because I want to reconstruct it into 3d and create counting system for how many bags in there. Thankyou sir I hope you can give your insight I'll subscribe to you
Not on its own. The differential renderer in the video is more of a tool that makes it possible to train machine learning models capable of reconstructing 3D objects using 2D training data.
Photogrammetry is a technique to reconstruct 3D from multiple 2D images. Meshroom is an open-source software that you could try today, but I doubt it would provide good enough results for what you need. But do you really need to reconstruct 3D from 2D?
What you are trying to do, sounds more like a 3D object detection problem. If I were you I would start with a paper like this one: arxiv.org/abs/2110.06922
Or, perhaps start simpler. You could also try to train an image classifier to count the number of flour bags based on on multiple images, without ever worrying about 3D.
But I am just speculating. This sounds like a research project.
@@CODEMENTAL Thankyou for the effort sir, very much appreciated 😇, God bless.
heeey where is cow what do you say???? I did not understand.....
But I have a challenge using TexturesUV instead of Texturesvertex
Can you provide more details on the issue?
@@CODEMENTAL Did you made UV Texture generation and 3D Model from front View Image or side view ?
:)
Stopping with your presentation grimaces in the foreground is ugly and counterproductive. For professional, serious and effective use, put a sober and simple presentation image.