Lu Lu

TL;DR: The proposed LPI-Net outperforms known advanced inpainting approaches with much fewer parameters and achieves an improvement of at least 3.52 dB of PSNR than other advanced approaches on CelebA dataset.

TL;DR: A novel convolution block is proposed to comprehensively capture the context information among feature representations and outperforms the current state-of-the-art inpainting approaches.

TL;DR: Experiments illustrate that the proposed distillation strategy and fusion module achieve better performance over the baseline technique, and the proposal outperforms the known state-of-art approaches in terms of single-stream and traditional two-stream methods.

TL;DR: A novel loss function containing the combination of edge guided loss term and weighted perceptual loss term was proposed and it was found that this method in objects removal task is visually plausible and pleasing which won the objects removal track in ICME 2019 challenge.