arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.14535cs.CVcs.LG
收藏

重新思考通过共享结构的编码器-解码器流程

Rethinking Encoder-Decoder Flow Through Shared Structures

Samsung R&D Institute UK (SRUK)(三星英国研发中心)

Frederik Laboyrie, Mehmet Kerim Yucel, Albert Saa-Garriga

AI总结:

针对密集预测任务中解码器结构不变的问题,提出共享结构“banks”为每个解码块提供额外上下文,通过重采样和特征融合提升基于Transformer架构的深度估计性能。

摘要:

密集预测任务的编码器架构日益复杂,然而解码器却基本保持不变。解码器依赖各个独立的块顺序解码中间特征图。我们引入了“banks”,即共享结构,每个解码块使用这些结构在解码过程中提供额外的上下文。通过重采样和特征融合应用这些结构,在自然图像和合成图像上,针对基于Transformer的最先进架构,在大型数据集上训练时,提升了深度估计的性能。

英文摘要:

Dense prediction tasks have enjoyed a growing complexity of encoder architectures, decoders, however, have remained largely the same. They rely on individual blocks decoding intermediate feature maps sequentially. We introduce banks, shared structures that are used by each decoding block to provide additional context in the decoding process. These structures, through applying them via resampling and feature fusion, improve performance on depth estimation for state-of-the-art transformer-based architectures on natural and synthetic images whilst training on large-scale datasets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑