Mar 28, 2024 · In this paper, we propose a novel auxiliary learning framework for mask-guided matting models, incorporating three auxiliary tasks: semantic segmentation, edge ...

Mar 28, 2024 · Second, we propose an inconsistency-guided detail regularization (IGDR) module to regularize detail refinement and avoid overfitting low-level ...

A novel auxiliary learning framework for mask-guided matting models is proposed, incorporating three auxiliary tasks: semantic segmentation, edge detection, ...

Image Matting is the process of accurately estimating the foreground object in images and videos. It is a very important technique in image and video editing ...

Learning Multiple Representations with Inconsistency-Guided Detail Regularization for Improving Mask-Guided Matting. To generalize mask-guided matting models ...

Learning Multiple Representations with Inconsistency-Guided Detail Regularization for Mask-Guided Matting. W Jiang, Z Xie, Y Lu, L Qi, J Cai, H Uchiyama, B Chen ...

We propose a new framework MaGGIe, Masked Guided Gradual Human Instance Matting, which predicts alpha mattes progressively for each human instances.

Figure 1 for Learning Multiple Representations with Inconsistency-Guided Detail Regularization for Mask-Guided. Abstract:Mask-guided matting networks have ...

本文旨在解决mask-guided matting模型在真实场景下存在的问题，例如过度拟合、缺乏对复杂结构的泛化能力以及受到背景线条或纹理的干扰。

May 8, 2024 · Learning Multiple Representations with Inconsistency-Guided Detail Regularization for Mask-Guided Matting. CoRR abs/2403.19213 (2024); 2023.