MAGIC: Mask-Guided Image Synthesis by Inverting a Quasi-Robust Classifier
Mozhdeh Rouhsedaghat, Masoud Monajatipoor, C.-C. Jay Kuo, Iacopo Masi

TL;DR
MAGIC introduces a mask-guided image synthesis method that inverts a quasi-robust classifier, enabling precise control over shape, location, and deformations in generated images with high fidelity and user control.
Contribution
The paper presents MAGIC, a novel one-shot mask-guided image synthesis technique that leverages structured gradients from a quasi-robust classifier for improved semantic preservation and control.
Findings
Achieves shape and location control in image synthesis.
Supports intense non-rigid shape deformations.
Provides firm user control via binary guide masks.
Abstract
We offer a method for one-shot mask-guided image synthesis that allows controlling manipulations of a single image by inverting a quasi-robust classifier equipped with strong regularizers. Our proposed method, entitled MAGIC, leverages structured gradients from a pre-trained quasi-robust classifier to better preserve the input semantics while preserving its classification accuracy, thereby guaranteeing credibility in the synthesis. Unlike current methods that use complex primitives to supervise the process or use attention maps as a weak supervisory signal, MAGIC aggregates gradients over the input, driven by a guide binary mask that enforces a strong, spatial prior. MAGIC implements a series of manipulations with a single framework achieving shape and location control, intense non-rigid shape deformations, and copy/move operations in the presence of repeating objects and gives users…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsImage Processing Techniques and Applications · Advanced Neural Network Applications · Cell Image Analysis Techniques
