MI-GAN 512 Places2 pipeline (ONNX) – mirror

Unmodified mirror of migan_pipeline_v2.onnx from https://huggingface.co/andraniksargsyan/migan (revision 406830d0fa60666da0071c342ad2fbc8f30c5c64), the ONNX pipeline published with MI-GAN: A Simple Baseline for Image Inpainting on Mobile Devices (Picsart AI Research, ICCV 2023). Code and weights are MIT licensed (see LICENSE; upstream https://github.com/Picsart-AI-Research/MI-GAN, LICENSE-WEIGHTS).

Inputs: image uint8 [1, 3, H, W] RGB; mask uint8 [1, 1, H, W] with 255 = keep, 0 = fill.
Output: result uint8 [1, 3, H, W].
sha256 6f1f3530a1a2324b19752018ce756088b07973cda8d7d890034ace5c8a48c40b.

Used by the techyCoderWorld photo editor, which runs it in the visitor's browser (ONNX Runtime Web).

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support