☆ 3.8 Proceedings Paper

Resolution-robust Large Mask In painting with Fourier Convolutions

2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022) (2022)

Related references

Note: Only part of the references are listed.

Article Computer Science, Artificial Intelligence

PEPSI: Fast and Lightweight Network for Image Inpainting

Yong-Goo Shin et al.

Summary: The PEPSI network utilizes a parallel decoding path architecture to improve image inpainting performance through both a coarse-to-fine inpainting path and a refined feature reconstruction path. Additionally, Diet-PEPSI significantly reduces network parameters while maintaining performance.

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness

Manyu Zhu et al.

Summary: In this study, a novel mask-aware inpainting solution is proposed by designing Mask-Aware Dynamic Filtering (MADF) module and adopting Point-wise Normalization (PN) to effectively learn multi-scale features for missing regions in the encoding and decoding phases. The end-to-end cascaded refinement model utilizes supervision information such as reconstruction loss, perceptual loss and total variation loss to improve inpainting results incrementally from coarse to fine, demonstrating effectiveness through extensive experiments on public datasets.

IEEE TRANSACTIONS ON IMAGE PROCESSING (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence