Total Variation
Total variation (TV) is a mathematical concept measuring the variation in a function's values, frequently applied as a regularizer in image and signal processing to promote smoothness while preserving sharp features. Current research focuses on extending TV's applications beyond traditional image denoising, encompassing diverse areas like biomedical imaging, tomographic reconstruction, and even machine learning, often employing advanced optimization techniques such as the alternating direction method of multipliers (ADMM) and incorporating variations like anisotropic and isotropic TV, or high-order TV. These advancements improve the accuracy and efficiency of various image and signal processing tasks, leading to better image reconstruction in medical imaging, enhanced denoising in hyperspectral imaging, and improved robustness in machine learning models.