State-of-the-art image reconstruction techniques for multi-channel inverse problems employ various inter-channel correlation and regularization techniques to address low data signal-to-noise ratio (SNR) and reconstruction artifacts. This can be imposed through statistical noise dependencies between channels or regularisation penalties employing joint structural regularities across the channel space. For the latter case one can use various vector and matrix norms enabling joint directions of smoothing. This mini-symposium will demonstrate current tendencies in multi-channel image reconstruction through correlative regularisation approaches as well as applications such as spectral computed tomography.