Context I am experimenting on this dataset various Deep Learning architectures. This dataset can be used in classification, localization, segmentation, generative models and face detection/recognition/classification. Content The dataset has 52156 rgb images. Train and test datasets are splitted for each 86 classes with ratio 0.8 . the original images has 1988x3056 dimension. It is reduced to 288x432 using OpenCV. Acknowledgements I download the books from different webpages. In the futures, I can add some new images if it needed. Inspiration Comic books have different images than standard images that are worked on. The characters, images, environments, colors, and more in this data set are much more challenging and confusing than the image data sets that have been worked on before. Besides, the results for the use of GANs are much bigger and more complicated than those that have achieved successful results.