论文标题
快速傅立叶内在网络
Fast Fourier Intrinsic Network
论文作者
论文摘要
我们解决将图像分解为反照率和阴影的问题。我们提出了在光谱域中运行的快速傅立叶固有网络FFI-NET，将输入分为几个光谱频段。 FFI-NET中的权重在光谱域中进行了优化，从而使收敛速度更快到较低的误差。 FFI-NET轻巧，不需要辅助网络进行培训。该网络是端对端训练的，具有新的光谱损失，该频谱损失衡量网络预测与相应地面真理之间的全球距离。 FFI-NET在MPI-SINTEL，MIT INTINCIC和IIW数据集上实现了最先进的性能。
We address the problem of decomposing an image into albedo and shading. We propose the Fast Fourier Intrinsic Network, FFI-Net in short, that operates in the spectral domain, splitting the input into several spectral bands. Weights in FFI-Net are optimized in the spectral domain, allowing faster convergence to a lower error. FFI-Net is lightweight and does not need auxiliary networks for training. The network is trained end-to-end with a novel spectral loss which measures the global distance between the network prediction and corresponding ground truth. FFI-Net achieves state-of-the-art performance on MPI-Sintel, MIT Intrinsic, and IIW datasets.
