Image restoration networks are usually comprised of an encoder and a decoder, responsible for aggregating image content from noisy, distorted data and to restore clean, undistorted images, respectively. Data aggregation as well as high-resolution image generation both usually come at the risk of involving aliases, i.e.~standard architectures put their ability to reconstruct the model input in jeopardy to reach high PSNR values on validation data. The price to be paid is low model robustness. In this work, we show that simply providing alias-free paths in state-of-the-art reconstruction transformers supports improved model robustness at low costs on the restoration performance. We do so by proposing BOA-Restormer, a transformer-based image restoration model that executes downsampling and upsampling operations partly in the frequency domain to ensure alias-free paths along the entire model while potentially preserving all relevant high-frequency information.

翻译：图像复原网络通常由编码器和解码器组成，分别负责从含噪、失真的数据中聚合图像内容，并恢复干净、无失真的图像。数据聚合与高分辨率图像生成通常都伴有混叠风险，即标准架构为在验证数据上达到高PSNR值，牺牲了重建模型输入的能力，其代价是模型鲁棒性低下。本研究表明，在最新重建Transformer中简单提供无混叠路径，便能在较低复原性能成本下支持模型鲁棒性的提升。为此，我们提出BOA-Restormer——一种基于Transformer的图像复原模型，该模型在频域内部分执行下采样和上采样操作，从而确保整个模型拥有无混叠路径，同时潜在保留所有相关高频信息。