4REIGN Studio — open-source notices Local AI uses MI-GAN, developed by Picsart AI Research. Code: https://github.com/Picsart-AI-Research/MI-GAN Weights: https://github.com/Picsart-AI-Research/MI-GAN/blob/main/LICENSE-WEIGHTS Pinned ONNX export: https://huggingface.co/andraniksargsyan/migan/tree/406830d0fa60666da0071c342ad2fbc8f30c5c64 Model: migan_pipeline_v2.onnx, 28,079,181 bytes SHA-256: 6f1f3530a1a2324b19752018ce756088b07973cda8d7d890034ace5c8a48c40b MI-GAN code license and weights license (both MIT): MIT License Copyright (c) 2024 Picsart AI Research (PAIR) Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. MI-GAN weights: MIT License Copyright (c) 2024 Picsart AI Research (PAIR) Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. ONNX Runtime Web 1.20.1 (Microsoft, MIT): MIT License Copyright (c) Microsoft Corporation Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. Local repair is an original, deterministic texture reconstruction implementation. It is a lower-detail fallback and does not use a neural model or network service. The browser downloads public model/runtime files. Photographs and masks remain on the device. No paid inference endpoint is used. Ordinary website hosting/bandwidth and the user's own device/data costs still apply. STUDIO 4 — THREE-ENGINE BACKGROUND REMOVAL Original — U²-Netp Model: BritishWerewolf/U-2-Netp, Apache License 2.0. Upstream: https://huggingface.co/BritishWerewolf/U-2-Netp Light — MODNet Model: onnx-community/modnet-webnn, Apache License 2.0. Upstream: https://huggingface.co/onnx-community/modnet-webnn Reference: https://github.com/ZHKKKe/MODNet Heavy — BiRefNet Lite (512×512) Model: studioludens/birefnet-lite-512, MIT License. Upstream: https://huggingface.co/studioludens/birefnet-lite-512 Authors: Copyright (c) 2024 ZhengPeng; Copyright (c) 2026 Studio Ludens V.O.F. MIT LICENSE (BiRefNet Lite and converted model weights) Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. ONNX Runtime Web — Microsoft Corporation, MIT License. https://github.com/microsoft/onnxruntime All three background-removal engines process photographs inside the browser. Model files and the runtime are fetched on demand; photo pixels are not uploaded by the model worker. Models are subject to browser memory and GPU support. HEIC/HEIF fallback decoder — heic2any 0.0.4 (loaded only when required) Copyright (c) 2020 Alex Corvi. MIT License. Source: https://github.com/alexcorvi/heic2any The browser loads the pinned unmodified library from jsDelivr; no image data is transmitted to the CDN. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.