Image credits & attribution
This page lists every upstream image, figure, and demo asset referenced in the AIFX documentation, with full attribution and the applicable license terms.
Policy
The documentation reproduces or links to selected figures and demo imagery from the upstream research projects that each plugin bridges to. This usage is for documentation, commentary, and educational purposes under applicable fair use / fair dealing provisions, with attribution to the original authors and citation to the source publications.
The repository’s BSD-3-Clause license applies to the plugin code only. Each image listed below retains the license of its original source. If you are an author and would prefer a different attribution form, or removal, please open an issue and we will respond promptly.
Sources by plugin
Each entry below lists every upstream asset referenced from the plugin’s documentation page (schema images, input/output demo videos and GIFs).
Depth Anything V3 (depth_da3)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page | Lin et al., ByteDance Seed, 2025 | Author copyright; reuse terms not explicitly stated | arXiv:2511.10647 |
assets/teaser.png (project page) — mirrored crop of its “Depth & Ray Maps” panel, self-hosted as assets/plugins/depth_da3/depth_da3_depth-map.jpg |
Same | Apache 2.0 attribution | Same |
| GitHub README assets | Same | Apache 2.0 (code repo) — README image reuse with NOTICE | Same |
| Hugging Face Space demo | Same | App: Apache 2.0; output: per generation | Same |
NormalCrafter (normal_crafter)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page | Bin, Hu, Wang, Chen, Wang, ICCV 2025 | Author copyright; reuse terms not explicitly stated | arXiv:2504.11427 |
img/pipeline.png and video/comparison/breakdance.mp4 (project page) |
Same | MIT attribution (project repo MIT-licensed) | Same |
| GitHub README assets | Same | MIT (code repo) — README image reuse with copyright preserved | Same |
DepthCrafter (depth_crafter)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page | Hu et al., Tencent AI Lab, CVPR 2025 Highlight | Tencent custom non-commercial license; documentation reuse with attribution | arXiv:2409.02095 |
img/overview.jpg, img/img01.jpg, img/d6_ours.png (project page) |
Same | Tencent non-commercial; fair-use citation | Same |
| GitHub README assets | Same | Same non-commercial; cite + attribute | Same |
SAM3 Segmentation (segmentation_sam3)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page | Carion et al., Meta AI Research, 2025 | © Meta Platforms; SAM License for code and weights | arXiv:2511.16719 |
assets/model_diagram.png and assets/dog.gif (GitHub repo) |
Same | SAM License — redistribution permitted with attribution | Same |
| Meta AI blog post | Same | © Meta Platforms; reuse with attribution per Meta’s standard terms | Same |
| GitHub README assets | Same | SAM License | Same |
MaMa Matting (matte_mama)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page | Lim, Oh, Huang, Yoon, Kim, Lee, CVPR 2026 | CC BY-NC 4.0 (code repo); reuse for documentation permitted with attribution | arXiv:2601.14255 |
assets/videomama.png, comparison/basket_rgb.mp4, comparison/basket_mask.mp4 (project page) |
Same | CC BY-NC 4.0; fair-use citation | Same |
| HF Space demo | Same | App: per HF Space terms | Same |
MatAnyone2 Matting (matte_ma2)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page (v2) | Yang, Zhou, Hao, Tao, CVPR 2026 Highlight | NTU S-Lab License 1.0 (non-commercial) | arXiv:2512.11782 |
assets/figures/matanyone1vs2.png and assets/videos_mat/mixkit-man-breakdancing-452-full-hd_78_{input,pha}_sm.mp4 (project page) |
Same | NTU S-Lab License 1.0; fair-use citation | Same |
| Project page (v1) | Yang, Zhou, Zhao, Tao, Loy, CVPR 2025 | Same | arXiv:2501.14677 |
SeedVR2 Upscaler (upscale_seedvr2)
| Source | Authors / Institution | License | Citation |
|---|---|---|---|
| Project page | Wang et al., ByteDance Seed, ICLR 2026 | Author copyright; reuse terms not explicitly stated | arXiv:2506.05301 |
images/result1.png (project page); demo MP4s under datasets/Iceclear/SeedVR_VideoDemos on Hugging Face |
Same | Apache 2.0 — redistributable with NOTICE preserved | Same |
| GitHub README assets | Same | Apache 2.0 — redistributable with NOTICE preserved | Same |
ComfyUI custom node attributions
Many plugins also reference ComfyUI custom node implementations by community authors. These wrappers do the bridge work between ComfyUI and the upstream model. The plugins are grateful to:
- PozzettiAndrea — ComfyUI-DepthAnythingV3, ComfyUI-SAM3
- AIWarper — ComfyUI-NormalCrafterWrapper
- akatz-ai — ComfyUI-DepthCrafter-Nodes
- okdalto — ComfyUI-VideoMaMa
- spiritform — comfy-matanyone2
- NumZ and AInVFX / Adrien Toupet — ComfyUI-SeedVR2_VideoUpscaler
Each is referenced from the relevant plugin documentation page.
OpenFX
The plugins implement the OpenFX specification, maintained by the Academy Software Foundation (openfx on GitHub). License: BSD-3-Clause.
Authors & funding
AIFX was developed by MaGMa for Reepost Studio, with funding from CNC (Centre national du cinéma et de l’image animée).