{"id":23673649,"url":"https://github.com/narcissusex/cunerf","last_synced_at":"2025-08-27T22:22:09.492Z","repository":{"id":187779640,"uuid":"667674830","full_name":"NarcissusEx/CuNeRF","owner":"NarcissusEx","description":"[ICCV2023] CuNeRF: Cube-Based Neural Radiance Field for Zero-Shot Medical Image Arbitrary-Scale Super Resolution","archived":false,"fork":false,"pushed_at":"2024-04-19T15:09:25.000Z","size":10130,"stargazers_count":23,"open_issues_count":1,"forks_count":0,"subscribers_count":3,"default_branch":"main","last_synced_at":"2024-04-19T16:30:24.485Z","etag":null,"topics":["iccv2023","medical-image-synthesis","neural-radiance-fields","pytorch","super-resolution","zero-shot-learning"],"latest_commit_sha":null,"homepage":"https://narcissusex.github.io/CuNeRF/","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/NarcissusEx.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null}},"created_at":"2023-07-18T04:02:47.000Z","updated_at":"2024-04-19T16:30:27.145Z","dependencies_parsed_at":null,"dependency_job_id":"f7a715ec-941f-4e05-a4ee-94f5b0a72b48","html_url":"https://github.com/NarcissusEx/CuNeRF","commit_stats":null,"previous_names":["narcissusex/cunerf"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NarcissusEx%2FCuNeRF","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NarcissusEx%2FCuNeRF/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NarcissusEx%2FCuNeRF/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NarcissusEx%2FCuNeRF/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/NarcissusEx","download_url":"https://codeload.github.com/NarcissusEx/CuNeRF/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":231737330,"owners_count":18418996,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["iccv2023","medical-image-synthesis","neural-radiance-fields","pytorch","super-resolution","zero-shot-learning"],"created_at":"2024-12-29T12:52:48.903Z","updated_at":"2024-12-29T12:52:49.499Z","avatar_url":"https://github.com/NarcissusEx.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# CuNeRF\nThe source code for our paper \"**[CuNeRF: Cube-Based Neural Radiance Field for Zero-Shot Medical Image Arbitrary-Scale Super Resolution](https://openaccess.thecvf.com/content/ICCV2023/papers/Chen_CuNeRF_Cube-Based_Neural_Radiance_Field_for_Zero-Shot_Medical_Image_Arbitrary-Scale_ICCV_2023_paper.pdf)**\", [Zixuan Chen](https://narcissusex.github.io), [Lingxiao Yang](https://zjjconan.github.io/), [Jian-Huang Lai](https://cse.sysu.edu.cn/content/2498), [Xiaohua Xie](https://cse.sysu.edu.cn/content/2478), *IEEE/CVF International Conference on Computer Vision* (**ICCV**), 2023.\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://narcissusex.github.io/CuNeRF/\"\u003eProject Page\u003c/a\u003e |\n  \u003ca href=\"https://openaccess.thecvf.com/content/ICCV2023/papers/Chen_CuNeRF_Cube-Based_Neural_Radiance_Field_for_Zero-Shot_Medical_Image_Arbitrary-Scale_ICCV_2023_paper.pdf\"\u003ePaper\u003c/a\u003e \n\u003c/p\u003e\n\u003cdiv align=center\u003e\n\u003cimg width=\"1148\" alt=\"framework\" src=\"assets/cunerf.png\"\u003e\n\u003c/div\u003e\n\n\n## Abstract\n\nMedical image arbitrary-scale super-resolution (MIASSR) has recently gained widespread attention, aiming to supersample medical volumes at arbitrary scales via a single model. However, existing MIASSR methods face two major limitations: \u003cb\u003e(i)\u003c/b\u003e reliance on high-resolution (HR) volumes and \u003cb\u003e(ii)\u003c/b\u003e limited generalization ability, which restricts their applications in various scenarios. To overcome these limitations, we propose Cube-based Neural Radiance Field (CuNeRF), a zero-shot MIASSR framework that is able to yield medical images at arbitrary scales and free viewpoints in a continuous domain. Unlike existing MISR methods that only fit the mapping between low-resolution (LR) and HR volumes, \u003cb\u003eCuNeRF\u003c/b\u003e focuses on building a continuous volumetric representation from each LR volume without the knowledge from the corresponding HR one. This is achieved by the proposed differentiable modules: cube-based sampling, isotropic volume rendering, and cube-based hierarchical rendering. Through extensive experiments on magnetic resource imaging (MRI) and computed tomography (CT) modalities, we demonstrate that \u003cb\u003eCuNeRF\u003c/b\u003e can synthesize high-quality SR medical images, which outperforms state-of-the-art MISR methods, achieving better visual verisimilitude and fewer objectionable artifacts. Compared to existing MISR methods, our \u003cb\u003eCuNeRF\u003c/b\u003e is more applicable in practice.\n\n\u003cdiv align=center\u003e\n\u003cimg width=\"1148\" alt=\"framework\" src=\"assets/framework.png\"\u003e\n\u003c/div\u003e\n\n## 1) Get start\n\n* Python 3.9.x\n* CUDA 11.1 or *higher*\n* NVIDIA RTX 3090\n* Torch 1.8.0 or *higher*\n\n**Create a python env using conda**\n```bash\nconda create -n cunerf python=3.9 -y\n```\n\n**Install the required libraries**\n```bash\nbash setup.sh\n```\n\n**[option] Install FFmpeg**\n```bash\napt install ffmpeg -y\n```\n\n\n## 2) Training CuNeRF for medical volumes\n```bash\npython run.py \u003cexpname\u003e --cfg \u003cconfig file\u003e --scale \u003cSR scale\u003e --mode train --file \u003cfilepath\u003e\n```\nSee *example_train.sh* for details, we also provide an example config file in the *configs* dir.\n\n## 3) Arbitrary rendering for medical slices\nRender slices at arbitrary positions (*zpos*: $-0.1$ ~ $0.1$), scales ($1$.x ~ $2$.x) and viewpoints (*angles*: $0$ ~ $360$ degrees) with an rotation axis $[1,1,0]$:\n```bash\npython run.py \u003cexpname\u003e --cfg \u003cconfig file\u003e --mode test --file \u003cfilepath\u003e --scales 1 2 --zpos -0.1 0.1 --angles 0 360 --axis 1 1 0 --asteps 45 \n```\nSee *example_test.sh* for details.\n\n## Citation\n\n```tex\n@InProceedings{Chen_2023_ICCV,\nauthor    = {Chen, Zixuan and Yang, Lingxiao and Lai, Jian-Huang and Xie, Xiaohua},\ntitle     = {CuNeRF: Cube-Based Neural Radiance Field for Zero-Shot Medical Image Arbitrary-Scale Super Resolution},\nbooktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},\nmonth     = {October},\nyear      = {2023},\npages     = {21185-21195}\n}\n```\n\n## Acknowledgement \n\nWe build our project based on **[NeRF-Pytorch](https://github.com/yenchenlin/nerf-pytorch)**. We thank them for their wonderful work and code release.","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnarcissusex%2Fcunerf","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnarcissusex%2Fcunerf","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnarcissusex%2Fcunerf/lists"}