Not specifically, more just the occasional paper you see posted where they've found a way to recover missing data from surrounding context. I.e. something like reverse engineering redaction boxes from JPEGs by reversing a non-lossless algorithm twice, once to get the lossy image in raw form with redactoin boxes, and at that point again to determine what was likely under those boxes from the surrounding lossy compression as it existed before.