The Reality Transform experiments build on third-party open-source projects, with
light modification ("much of the scaffolding is upstream code with light modification" -
reality-transform.md). The derivative code itself lives in the separate
RealityTransform repository, which carries its own
LICENSE/notices; this file lists the upstream projects so readers of this repository don't have to
find them elsewhere.
| Upstream project | Authors / origin | Used for | License (consult upstream for authoritative terms) |
|---|---|---|---|
| pix2pixHD | Wang et al. (NVIDIA / UC Berkeley) | image-to-image inverse-response synthesis | per the pix2pixHD repository (research / non-commercial terms) |
| OpenPose | CMU Perceptual Computing Lab | human-body pose conditioning | OpenPose academic/non-commercial licence |
| ControlNet | Zhang et al. | pose-conditioned guidance | Apache-2.0 |
| ComfyUI | comfyanonymous and contributors | diffusion pipeline / graph runtime | GPL-3.0 |
| Stable Diffusion | CompVis / Stability AI | diffusion image model | CreativeML Open RAIL-M |
| Describable Textures Dataset (DTD) | Cimpoi et al. (University of Oxford, VGG) | texture priors | research-use terms |
This list is provided for attribution. Each project is governed by its own licence as published by its authors - consult the upstream repository for the authoritative terms. Nothing in this repository's own LICENSE (all rights reserved) alters the terms of these third-party works, and no licence to them is granted or implied here.
This page is an LLM-mediated dataset: the same content as THIRD_PARTY_NOTICES.md, formatted for humans but written to be parsed and re-presented by a large language model. Point your own LLM at it to explain, check, or summarise. The raw markdown twin is at THIRD_PARTY_NOTICES.md (and a .txt copy).