temple_buddhist Project Tirtha Beta

About Kadamba Shri Mahadeva Temple

Archaeological site in a tranquil area consisting of a 12th-century Hindu temple carved from stone. (text from Google Maps). Location: https://maps.app.goo.gl/tYs4jvx1GdkM6P757

Model Statistics

What is Project Tirtha?

Project Tirtha is an academic initiative to create 3D models of heritage sites, using crowdsourced images. The word Tirtha is Sanskrit for "a place of pilgrimage", and is commonly used to refer to the sacred sites of Hinduism, Jainism and Buddhism. Our goal is to preserve and showcase the beauty and cultural significance of heritage sites. We believe that by allowing the general public to contribute to the creation of these models, and by providing open access to these models, we can increase awareness and appreciation of these important cultural landmarks and inspire future generations to maintain them for years to come.

This project is open-source under the GNU Affero General Public License v3.0 and is under active development. You can find the source code on our GitHub repo. All contributions are welcome. Please read CONTRIBUTING for more details. You can report bugs or suggest features, submit images via "Contribute", or suggest heritage sites via "Request site". Please read How do I contribute? before proceeding. A slideshow with video instructions for the project can be found here, while a slidedeck presented at ACM Web3D 2023 can be found here.

Please cite the following paper if you use this software in your work:

@inproceedings{10.1145/3611314.3615904,
    author = {Shivottam, Jyotirmaya and Mishra, Subhankar},
    title = {Tirtha - An Automated Platform to Crowdsource Images and Create 3D Models of Heritage Sites},
    year = {2023},
    isbn = {9798400703249},
    publisher = {Association for Computing Machinery},
    address = {New York, NY, USA},
    url = {https://doi.org/10.1145/3611314.3615904},
    doi = {10.1145/3611314.3615904},
    booktitle = {Proceedings of the 28th International ACM Conference on 3D Web Technology},
    articleno = {11},
    numpages = {15},
    keywords = {photogrammetry, open source, digital heritage, crowdsourcing, 3D dataset},
    location = {San Sebastian, Spain},
    series = {Web3D '23}
}
            

How do I contribute?

Upload instructions

Before you begin, please read the following instructions:

  • Ensure you have a Google account for authentication.
  • Click "Contribute" to open the upload form.
  • Read the Terms of Use and Privacy Policy and click on the checkbox, if you agree.
  • Authenticate by clicking "Sign in with Google". Only your email address and profile name will be accessed.
  • After authentication, the upload form will be enabled. Select the heritage site and upload images. The list of available sites is on the right. "Open" means contributions are being accepted, "Closed" means they are not.
  • Upload up to 2000 images (total size of < 1 GB) that meet these requirements for the internal photogrammetry pipeline:
    • The object of interest (a part of the building) is clearly visible, without obstructions (people, vegetation, etc.), and in focus.
    • Images are taken from multiple angles or viewpoints with 30 - 40% overlap between images.
    • Lighting conditions are consistent with no dark shadows, reflections, or transparent objects. Prefer diffused lighting.
    • Images have minimal blur, noise, distortion, or pixelation.
    • Use either landscape or portrait orientation (do not mix both) to avoid distortion.
    • Images must be above 1080p in resolution and contain EXIF data, mainly the camera parameters.
    • In general, try to upload images for a low-resolution or missing part of the building in the live model.
    • Try to be as safe as possible while taking images. Do not trespass or put yourself in danger.
    Images not meeting these requirements will be filtered out (as they may not contribute to the model). Confirm that you are not uploading images that[Check ToS & Privacy Policy]:
    • contain sensitive information or personal data.
    • are not yours or you don't have the right to upload.
    • are unrelated to the site you're uploading to.
    All uploads are licensed under CC BY-NC-SA 4.0 as specified in the Terms of Use.
  • On desktop, you can drag and drop multiple images at once to the upload area. On mobile, you can select multiple images from your gallery.
  • Upon successful upload, a confirmation message will appear, and the form will reset. You can close the form or upload more images.
  • Note: It may take a few days to process the images and update the models.

Troubleshooting

If you encounter errors, try the following:

  • Enable cookies & JavaScript in your browser if you can't upload images.
  • Ensure your ad-blocker isn't blocking necessary scripts. No ads are served on this site.
  • Try signing in again if you get an authentication error.
  • If the error persists, hard-refresh the page (Ctrl + Shift + R on Windows/Linux, Cmd + Shift + R on Mac), and try uploading again. If issues persist, open an issue on GitHub.
How does it work?

This project utilizes open-source libraries and automated pipelines for photogrammetry (based on AliceVision) and 3D Gaussian Splatting to create 3D models from crowdsourced images of heritage sites. On the photogrammetry side, we broadly perform the steps described at AliceVision | Photogrammetry Pipeline. The generated textured mesh is denoised, decimated, and converted to a .glb file using obj2gltf. This file is optimized for web use with meshoptimizer. Finally, the 3D model is rendered in the browser using <model-viewer>.

For 3D Gaussian Splatting, we use the splatfacto implementation present in the excellent nerfstudio, which produces a Gaussian point cloud. This is then compressed and filtered and converted to .splat file for viewing on the web using a WebGL-based 3D Gaussian Splat Viewer. The fuzz / floaters present in the Gaussian Splat models are artifacts of the splatting process and are not part of the original model.

Please note that the models displayed here are low-poly, compressed versions due to web & mobile device constraints.

Acknowledgment

We thank the following individuals for their contributions to the project's development:

We are grateful to the developers of the following open-source libraries, which help make this project a reality:

We also thank Odisha State Archaeology for their support.

Terms of Use [Updated: July 15, 2023]

[Effective: July 15, 2023] By accessing this website or uploading images to Project Tirtha ("use"), you agree to be legally and contractually bound by these terms of use and our Privacy Policy. If you do not agree to these terms or our privacy policy, do not access or use Tirtha.

  • Tirtha is an academic image crowdsourcing platform specializing in heritage sites, created by Dr. Subhankar Mishra's Lab at National Institute of Science Education and Research. Uploaded images are used to create 3D models through a photogrammetry pipeline. By using Tirtha, you are contributing to this process.
  • By using Tirtha, you warrant that you have legal rights to share the uploaded images, which may be used by us for creating 3D models. All images uploaded to Tirtha are licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
  • We reserve the right to remove inappropriate user content, including images unrelated to heritage sites, personally identifiable information, and any inappropriate content.
  • We reserve the right to terminate your use of our platform for repeatedly uploading inappropriate or unrelated images, personally identifiable information, or any inappropriate content. Termination prevents further image uploads but allows access to view 3D models. We will make our best efforts to notify you of termination or if our systems flag any such content, but we are not liable for any data loss or damage that may occur as a result of such termination or flagging.
  • The accuracy of 3D models depends on various factors, and we strive for accuracy, but we are not responsible for inaccuracies, errors in models, or any damage or data loss during the image upload process.
  • The outputs, which include the 3D models created from images submitted to Tirtha, are licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Please check if your intended use aligns with this license, and contact us if unsure.
  • We reserve the right to change the license on the outputs, including the 3D models, at any time, which will affect models reconstructed after the date of the license update. Check the model information in the About section for the reconstruction date.
  • By using Tirtha, you agree to indemnify and hold us harmless from any claims, damages and expenses (including attorneys' fees) arising from your use of our platform. This also includes any claims, damages, or expenses arising during the image capture process. It is the sole responsibility of the user to ensure that the image capture process is safe and legal.
  • We reserve the right to change these terms at any time. Material changes will be notified via email or via a message posted on the site. Your continued use of Tirtha after the effective date of the modified terms constitutes acceptance of the modified terms.
  • These terms of use are governed by the laws of India, and any dispute will be subject to the exclusive jurisdiction of the courts of Bhubaneswar, Odisha, India.
Privacy Policy [Updated: September 18, 2024]

[Effective: September 18, 2024] Please read our privacy policy carefully before using Tirtha. Do not access or use Tirtha if you do not agree with any part of this policy.

  • We use Google OAuth2 for authentication of contributors, i.e., "people who upload images". This is done via Authlib's Django OAuth Client and no tracking cookies are used. Only your email address and public profile name ("personal information") are stored to identify your contributions, i.e., "sets of images uploaded by you", on this website.
  • Uploaded images are stored on our servers, and the associated metadata, including EXIF data and GPS coordinates (if available), is used solely for 3D model reconstruction purposes.
  • All pre-processing steps, except NSFW filtering, occur client-side on your browser or server-side on our end. We utilize local models and the Google CloudVision API for NSFW image filtering. The images are uploaded to Google servers for filtering purposes but are not stored on them.
  • Your personal information and uploaded images are not sold or rented to any third party.
  • Session cookies are utilized to store your login information, and they are deleted when you close your browser. We do not use other cookies or analytics tools on this website.
  • Your personal information is retained as long as you have at least one contribution associated with your account. If all contributions are deleted, your personal information and account are also deleted within a reasonable amount of time, usually less than a week. You can create a new account and upload new images by reauthenticating.
  • To request the deletion of your account and personal information, or some/all of your contributions, please contact us at project.tirtha@niser.ac.in. Please note that not all contributions may be deleted if they were used to create 3D models.
  • We may use your personal information to contact you about the project or your contributions. We may also share your information if required by law.
  • Appropriate technical and organizational measures have been implemented to protect data on our servers. However, we cannot & do not guarantee the absolute security of your information or uploaded content.
  • Tirtha is an academic project, and we disclaim liability for any data loss or damage resulting from its use. Your use of this website is at your own risk, and you agree to indemnify us against any claims arising from such use.
  • We reserve the right to modify this privacy policy at any time. Please refer to the "Effective" date above for the last update.
fullscreen

add_business Request site

Contribute to Tirtha

Please sign in to upload images.
No files selected.

All contributions licensed under CC BY-NC-SA 4.0.