The field of image processing is moving towards more sophisticated and specialized techniques for enhancing and reconstructing images in various applications, including autonomous driving, endoscopy, and multimode fiber imaging. Recent developments have focused on addressing challenges such as low-light conditions, noise, and blur, with a emphasis on unsupervised and self-supervised learning methods. Noteworthy papers in this area include the Unified Low-Light Traffic Image Enhancement via Multi-Stage Illumination Recovery and Adaptive Noise Suppression, which proposes a fully unsupervised multi-stage deep learning framework for low-light traffic image enhancement. The Nested Unfolding Network for Real-World Concealed Object Segmentation is also notable for its unified framework for real-world concealed object segmentation. Additionally, the HistoSpeckle-Net: Mutual Information-Guided Deep Learning for high-fidelity reconstruction of complex OrganAMNIST images via perturbed Multimode Fibers proposes a deep learning architecture for reconstructing structurally rich medical images from multimode fiber speckles.