iColoriT: Towards Propagating Local Hint to the Right Region in Interactive Colorization by Leveraging Vision Transformer

14 Jul 2022  ·  Jooyeol Yun, Sanghyeon Lee, Minho Park, Jaegul Choo ·

Point-interactive image colorization aims to colorize grayscale images when a user provides the colors for specific locations. It is essential for point-interactive colorization methods to appropriately propagate user-provided colors (i.e., user hints) in the entire image to obtain a reasonably colorized image with minimal user effort. However, existing approaches often produce partially colorized results due to the inefficient design of stacking convolutional layers to propagate hints to distant relevant regions. To address this problem, we present iColoriT, a novel point-interactive colorization Vision Transformer capable of propagating user hints to relevant regions, leveraging the global receptive field of Transformers. The self-attention mechanism of Transformers enables iColoriT to selectively colorize relevant regions with only a few local hints. Our approach colorizes images in real-time by utilizing pixel shuffling, an efficient upsampling technique that replaces the decoder architecture. Also, in order to mitigate the artifacts caused by pixel shuffling with large upsampling ratios, we present the local stabilizing layer. Extensive quantitative and qualitative results demonstrate that our approach highly outperforms existing methods for point-interactive colorization, producing accurately colorized images with a user's minimal effort. Official codes are available at https://pmh9960.github.io/research/iColoriT

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Point-interactive Image Colorization CUB-200-2011 iColoriT PSNR@1 27.986 # 1
PSNR@10 30.595 # 1
PSNR@100 33.543 # 1
Point-interactive Image Colorization CUB-200-2011 iDeepColor PSNR@1 27.45 # 4
PSNR@10 29.32 # 4
PSNR@100 31.57 # 2
Point-interactive Image Colorization CUB-200-2011 SWF PSNR@1 23.547 # 6
PSNR@10 25.097 # 6
PSNR@100 27.623 # 6
Point-interactive Image Colorization CUB-200-2011 InstColor PSNR@1 27.69 # 2
PSNR@10 29.45 # 2
PSNR@100 31.45 # 4
Point-interactive Image Colorization ImageNet ctest10k iColoriT PSNR@10 30.626 # 1
PSNR@1 27.474 # 1
PSNR@100 33.787 # 1
Point-interactive Image Colorization Oxford 102 Flowers iDeepColor PSNR@1 22.72 # 4
PSNR@10 25.13 # 2
PSNR@100 27.826 # 2
Point-interactive Image Colorization Oxford 102 Flowers SWF PSNR@1 18.452 # 6
PSNR@10 19.445 # 5
PSNR@100 22.362 # 6
Point-interactive Image Colorization Oxford 102 Flowers InstColor PSNR@1 22.97 # 1
PSNR@100 27.35 # 4
Point-interactive Image Colorization Oxford 102 Flowers iColoriT PSNR@1 22.925 # 3
PSNR@10 27.37 # 1
PSNR@100 30.731 # 1

Methods