Revisiting Context Aggregation for Image Matting
This addresses performance degradation in image matting due to context scale shifts, offering a novel approach that improves accuracy for applications like photo editing, though it is incremental in refining network design.
The paper tackled the problem of context scale shift in image matting by showing that a basic encoder-decoder network without specialized context aggregation modules can learn more universal context aggregation, leading to higher performance. It introduced AEMatter, which outperformed state-of-the-art methods by a large margin on five datasets.
Traditional studies emphasize the significance of context information in improving matting performance. Consequently, deep learning-based matting methods delve into designing pooling or affinity-based context aggregation modules to achieve superior results. However, these modules cannot well handle the context scale shift caused by the difference in image size during training and inference, resulting in matting performance degradation. In this paper, we revisit the context aggregation mechanisms of matting networks and find that a basic encoder-decoder network without any context aggregation modules can actually learn more universal context aggregation, thereby achieving higher matting performance compared to existing methods. Building on this insight, we present AEMatter, a matting network that is straightforward yet very effective. AEMatter adopts a Hybrid-Transformer backbone with appearance-enhanced axis-wise learning (AEAL) blocks to build a basic network with strong context aggregation learning capability. Furthermore, AEMatter leverages a large image training strategy to assist the network in learning context aggregation from data. Extensive experiments on five popular matting datasets demonstrate that the proposed AEMatter outperforms state-of-the-art matting methods by a large margin.